The FanGraphs Auction Calculator!
I’m pleased to announce that FanGraphs now has an auction calculator!
In terms of methodology, we are using a method that is pretty similar to the Last Player Picked methodology.
Besides all the usual features of an auction calculator, we have a couple of unique features that are worth some addition explanation:
– There is an “experimental” check box, which adds in replacement level value for players with incomplete playing time. If a player was to have a .300 batting average for the season in 450 plate appearances, it would add in an extra 150 plate appearances at a .220 average (numbers for illustration only) for that player and calculate the auction values off his new batting average.
This is something that could be useful if your league allows you to carry a lot of bench spots and you have the ability to supplement potentially high risk injury players with a replacement level player. Chad Young wrote about this a few years ago.
Please feel free to check the box and see how it changes player values if you feel it’s appropriate for your league. And like the option says, this is still an experiment and should be considered a “beta” feature.
– There is also an option to artificially increase or decrease relief pitcher value. Since we don’t project holds, this might be particularly useful for ottoneu leagues or other leagues where holds are considered. By not taking into account holds for a league that might consider them, relief pitcher values will most likely be inflated.
This could also be a useful option if the players in your league generally undervalue (or over value) relievers and you would like to put relief pitcher values more in line with reality as opposed to theoretical value.
Please let us know if you run into any issues!
David Appelman is the creator of FanGraphs.
Great Stuff! I understand that it is difficult and probably won’t happen, but the ability to use your own custom projections would be fantastic.
Please add the ability to use playing time projections from your depth charts (instead of Steamer). The Steamer playing time projections have not been anywhere near as accurate historically.
Also, are Oliver projections coming this year?
Oliver projections are coming soon, and Steamer already uses the playing time projections on the depth chart.
Steamer is def. *not* synched to your depth charts. E.g., as of the morning of Feb 25, Steamer shows 677 projected PA for Mike Trout:
http://www.fangraphs.com/statss.aspx?playerid=10155&position=OF
Vs. 644 PA on your Depth Charts:
http://www.fangraphs.com/depthcharts.aspx?position=ALL&teamid=1#ALL
David, Is there any update on this issue? None of the playing time projections match between Steamer and your depth charts.
the position adjustment for DH appears to be positive instead of negative
nevermind, everything is backwards in the list I export. Trout is worth negative $119
Are you seeing a problem with the export where it exports things in reverse or is everything ok?
Yes, J. Giambi is the most valuable hitter and Trout the least.
I’m not seeing this when I export. Can you post a link to the page you are trying to export from?
http://www.fangraphs.com/auctiontool.aspx?type=bat&proj=steamer&pos=1,1,1,1,4,1,0,0,1,1,0,0,8,5&dollars=700&teams=20&mp=10&msp=5&mrp=5&mb=35&split=&points=c|0,1,2,3,4,5,6,17|0,1,2,3,4,8&lg=MLB&rep=0&drp=0&players=
I think it’s because the $35 min price x 25 players is > than the 700 cap. You can disregard, this tool won’t work for our league because we auction part of it until people are out of money and then draft. Sorry for the confusion. Thanks.
Yeah, I after looking, I was just going to say that a $35 min doesn’t seem possible in your league. You might getting rid of your bench entirely because there’s no way you would bid on bench players in your league since they would go straight to draft.
Interesting setup.
David, I’m also seeing issues with DH. For instance, David Ortiz projects as $1 using FGPoints Linear Weights for Ottoneu. In addition, adjusting the eligibility toggles (Ortiz started 5 games at 1B in 2014), does not seem to change his position eligibility to 1B instead of DH.
Oh, actually it looks like the positional eligibility is set as > instead of >= to. I’ve changed this and David Ortiz should show up as position eligible for 5 games now. He would have shown up at 1B if you put 4 Games previously.
Thanks! Any update on when the full ZiPS projections will be uploaded?
Can you copy and paste into the keeper exclusion list?
If you have a list of playerids, you can add them as playerid|value,playerid|value to the players= in the url.
Other than that, there’s currently no way to do this.
Great work!
Just curious, is there any way to add in stats? For example, my league uses Quality Starts as a category.
Thanks!
I second Quality Starts!
3rd Quality Starts!
Love it. Also agree on needing quality starts
Absolutely yes for quality starts. That would make a huge difference for me.
QS please!
Innings Pitched would be nice as well
2nded IP stat category request
QS and IP are needed indeed.
Otherwise awesome tool!
Is catcher pos. adjustment too high? Mcanns price seems out of whack
It’s going to depend on your player pool and the projections. If you reduce the number of starting catchers in your league to 1, their value will decrease considerably.
great stuff!
is there a way to save the settings, like in the FG+ custom teams tab? I’m guessing many of us are in multiple leagues with different settings and keepers and it would be helpful to be able to save that info.
We will most likely add this in future update, but for now you will have to just bookmark the page you want to save. Whatever the url is, it will take you back to the same page with all your settings saved.
Wait – when does the Pauction Nalculator come out, then?
Awesome, any change Holds can get added as a pitching field?
Thanks!
Yeah, I was coming on to post the same thing. Either Holds or Saves+Holds would be very valuable.
The issue right now is we do not have holds projected, so there’s not a whole lot we can calculate without that.
As a quick adjustment, you might want to try the “artificially decrease relief pitcher” option.
We’ll see what we can do though. I would definitely like to include holds and holds + saves as categories.
Thanks for the quick response, and great work on this, it’s amazing!
how much should we decrease/increase?
Terrific. You might edit the comment for batting / pitching split to more clearly indicate that the user should enter he percentage (omitting the % sign) of dollars to be allocated to batters. I figured it out after putting a dollar amount there first.
Please add HBP and Holds to the calculator so those of us in FanGraphs points leagues can make full use of it!
I assume quality starts is omitted because it’s not a projected stat yet. Nevertheless, very cool stuff. Thanks, FanGraphs!
Great work!
Just curious, is there any way to add in stats? For example, my league uses Quality Starts and extra points for extra base hits as a category.
Thanks!
Great work! As far as ottoneu is concerned, I have built an auction calculator that accounts for historical ottoneu replacement levels. It allows you to import your own projections as well. If anyone is interested, let me know.
Hey ottoneutrades I would love that calculator. I think I wound up with a version you made in the past as well, great stuff.
Fun drop offs in my league:
$15 from Tulo to HanRam.
Very cool, but some leagues use Runs Produced (R+RBI-HR), you should add that in.
Anyway to make it stop over-rating relievers compared to starters a.k.a. balance by playing time contribution? Unless you think a guy like Julio Teheran goes for -$2.30 in a normal 5×5 with a sixth cat for K:BB. The batter stuff is really cool at least and seems to be working fine
I think Dave mentioned this in the post, but there is a small option near the bottom of the inputs part of the page where you can “devalue” relief pitchers by an amount you select (%) across the board. This way, you can make relievers’ values better fall in line with starters.
Dubious Stat Request #1:
I see you have RBI + R already, which I can use in a pinch. But my league uses RBI + R – HR. Any chance you can add that as well?
Dubious Stat Request #2:
Slightly harder to get around, but since I assume singles are projected, could you add 1B to go along with 2B and 3B?
Either way, thanks so much for the awesome tool!
Oh, and one more thing — any chance you could use the tool’s algorithm to show dollar values for a custom date range in the past? So if we want to see what players actually earned in 2014, we could. Or even if we wanted to see what players earned across multiple years, or a portion of the current season to date?
Ideally you could calculate total cumulative dollar values, and some sort of per-game dollar value to pick up players whose total value would be otherwise decreased by injury.
Oops, I see the RBI + R – HR request was already made above. But at least now I can take comfort that my league isn’t the only league that uses it.
But I really decided to leave another comment to thank you for something I didn’t notice the first time — that you allow us to enter keepers. That’s going above and beyond what most sites do, and an example of what makes FanGraphs great.
Unless we’re both using weird pseudonyms and are in the same league. 😉
Highlight and Google: “Tinfoil Hats”
What-the-what? I was responding to some crazed wingnut who commented, but then it dumped me into the main comment thread (I assume because the original comment was deleted).
Haven’t seen that behavior here before. Looks like the new comment system is a bit weird.
I would like to add IP to the list of categories.
Couple of questions when it comes to keepers. I see it said to remove a keeper, put in their name and leave price blank. That didn’t remove the player from the list for me. When I put in my league’s keeper cost for that player, it did add it to the list in a new column. Will it recalculate other players’ values to account for the removed player? Or just for ease? Thanks.
You need to add him as a keeper with some value (even if it’s $0). Then if you add a keeper and later want to remove him as a keeper you can just leave the price blank.
There is no way to completely remove a player from the list of players because regardless of whether the player is on a team or not, as long as he is in the player pool, you do need to calculate his value to calculate everyone else’s value.
Just to follow up, and make sure I’ve got this. Lets say I got stupidly lucky his Rookie year and have Mike Trout for $12 this year.
If I enter him as a keeper @12 will it 1) will it allocate $12 from the $$ pool to Trout, and 2) Thereby increase the value (ever so slightly) of all the other players as there is more money to spend on them with player X out of the running?
I will likely say it again, but great great work!!! love it in its current form and reading the comments it looks like it is only going to be getting better.
This is fantastic. When all the Zips projections are up at FG will we be able to create the values based on those projections as well? Also, just out of curiosity, is there any chance more pitching cats will be added? The list for hitting cats is expansive, but pitching seems to fall a bit short.
Thanks for this, it’s a great tool.
I’m about halfway done entering keeper values for a league and can’t get any further because the site keeps timing out. Curious if anyone else is having similar problems? Is there a limit to the number of keepers you can enter? I got to about 200 before I started having issues.
If you e-mail me the very long url (david at fangraphs.com), I’ll take a look. There might be an url length limit, or some limitation on our end.
Thanks, this is great. Two requests:
– the ability to differentiate by OF type (LF/CF/RF)
– the ability to use a blended Steamer + ZIPS once the final zips is available. (Having that data anywhere on the site would be most helpful).
The Steamer projections cannot guess on playing time, however you do guess the playing time on your depth charts pages … why don’t you adjust the Steamer stats based on those more realistic PAs?
Thanks!
The Steamer projections currently do use the depth charts to determine playing time. The playing time totals are going to be nearly identical between the two.
Oh that is good thanks
Why no QS?
so how much are people actually paying for Gattis in a 10 team, 6×6 roto league (2C) ? *assuming you agree with the projection
This lists $26.7
This is awesome, thanks guys. Quality starts (if it gets projected) and total bases added to the points league section would be nice.
Thanks Fangraphs! Now all you need to do is offer MockDrafts! Seriously, I don’t know a better way to bring in traffic!
You could make MockDrafts a part of Fangraphs+
Can you give the ability to combine multiple projection systems, like Steamer, ZiPS and fans? That would be huge
Can you give the ability to combine multiple projection systems, like Steamer, ZiPS and fans? That would be huge.
Maybe I’m just being a bit dumb here, but when adding the points and position adjustment columns together, they come out to a $1 less than the total end column.
When calculating $ values, you add the minimum bid amount to each player after you are done calculating the values for each category, so each player should be worth the minimum bid more than the sum of their components.
Got it. Thanks.
Pitcher home runs allowed as a stat.
I loved this until I saw Todd Frazier valued at $1.1. Then I was sad.
don’t hate the fangraphs. hate the steamer
Or hate the Todd Frazier.
Currently, a starter and reliever with the same ERA seem to be given the same amount of credit toward ERA. The starter is expected to throw more innings, and will therefore have a greater impact on your team ERA.
You can account for this by converting all categories to rate stats, normalizing, then multiplying the end result by the number of PA or IP. This also better accounts for the value of platoon players or partial seasons.
The way you describe it should be working is currently the way it is working. We do take playing time into account with all rate stats. You can see this in a setup where ERA is the only stat:
ERA Only Category Pitchers
If we weighted them the same regardless of innings, Chapman would be the #1 pitcher by far. However, his 65 innings of 1.49 ERA is worth $14.6 compared to Kershaw’s 201 innings of 2.42 ERA worth $20.0
Correct. Great application, this saves a lot of work for my auction prep.
What is AVG in the pitching categories?
Thanks
Adding keepers only seems to adjust the values up to a point then they seem to be static. I noticed my players weren’t changing value at all after they gained about 12% value, even with hundreds of dollars additional keeper savings added in.
If you give me a link to see what you’re talking about, I can see if there’s a problem or not, but you can get the values up a lot higher than 12%:
Crazy Inflation Link
here’s the link
http://www.fangraphs.com/auctiontool.aspx?type=pit&proj=steamer&pos=2,1,1,1,5,1,1,1,0,2,2,2,5,6&dollars=260&teams=15&mp=20&msp=5&mrp=5&mb=1&split=69&points=c|0,1,2,3,4|0,1,2,3,4&lg=MLB&rep=0&drp=0&players=5038|4,9166|32,6893|9,2429|5,4949|51,4881|11,1887|23,13431|2,4106|5,14078|11,2434|8,14225|8,15676|6,4772|28,10954|2,7859|2,6632|2,12861|8,8203|8,4930|9,9218|29,745|9,9784|2,9810|10,7196|10,2151|33,4022|4,6184|2,9241|4,6562|3,10324|2,5524|24,11713|2,10762|2,5666|6,785|3,12856|3,6885|23,11530|3,2036|1,10603|1,3137|1,10233|1,10131|1,1943|1,3184|1
The inflation values here look correct. You are going to have to enter in a lot more players at very low prices for things to change drastically.
ok, i did a test and saw something interesting. Even adding another 15-20 of top pitchers in @$1 didn’t change the price of the pitchers by a cent.
However, adding in even one top hitter at a dollar value of $1 changed the prices of the top guys by .40 to .60 for every one i entered.
Not sure why but the pitchers stopped adjusted value after about 10 were entered in.
I’ll try starting over, just wanted to let you know of a possible bug incase someone else notices it too. Very cool tool, though, great works guys!
I should also add that in a league with real people and not theoretical teams, you’re not going to see completely absurd inflation. 12% seems fairly reasonable.
This is great. Thanks again, David. Any chance you could add HR/9 and Innings to the pitcher stats?
Seconded. My ottoneu league uses HR/9.
Is there any way to display the actual projected STAT numbers for each category (in addition to the total projected dollar value), instead of/in addition to the category weighted dollar amount?
Steamer stats are available literally 1 or two clicks away, dude.
Some of these comments are outrageous in their entitlement.
I’m not asking they bend over backwards nor am I asking for something difficult to display. I’m asking because if I view the page and/or export the data via Excel it would be nice to see the projected stats in addition to the projected value of which they are based upon. Sorry for my entitled and outrageous request . . . smh
Google VLOOKUP.
Whats criminal is the amount of work invloved in proving im not a robot before posting a comment.
Wow! This is absolutely awesome! Last year I spent days looking for something like this, this is above and beyond what I was hoping I could find!
Ok, to summarize and suggest:
1) Excellent work. Great job! Not that I needed another reason, but you added a big one. Fangraphs is where it is at.
2)IP would be awesome to be added as a stat category
3)Please see earlier question about Keeper operation.
4)Fantastic Job…Just Great all around. Just Great.
5)Suggest that when you export the results into an excel file, one tab be hitters, and one be pitchers (I know I can do this on my own, and have, but just wanted to mention it as an idea).
6)I like the idea of blending projections, but I think something even better would be a report that lists both values by each player, thus allowing me to manipulate the data myself (floor, ceiling, average, whatever) without having to port one set of data into a different spreadsheet.
7) Did I say great job yet, because Great Job! Thank you Fangraphs.
Thanks! As far as keepers go, in your above comment, you are 100% correct in describing how it works.
We will probably have a combined tab for all players, and we’ll be adding more projections as they become available. I’m not sure we’ll be able to do #6 though.
This is great; but, could you please at least add some built in options for Sabr and ottoneu points so I don’t have to put in what each value is worth every time I care to do one of these?
Any settings you enter can be bookmarked in your browser, or wherever until we get some preset options. That should be done pretty quickly.
Hey, can you make the league setting more customizable? I think on Last Player Picked, you were able to select specific teams. I only ask (and maybe this is only my backwards league) but my NL only dynasty league didn’t know how to deal with the astros leaving the NL, especially since we had a lot of prospects, so our 13 team league is NL+the Astros, for example.
Thanks!
I like the experimental feature to add in replacement player value. Where do the replacement player stats come from? I do the same thing manually and I take an average of the 20 worst players that are in the positive value pool and the 20 best players not in the positive value pool (So 40 Players Total). I don’t factor in position at all. Do the replacement stats consider position?? Like, catcher would have way worse replacement stats than 1Bs?
One request for additional flexibility is to only “top up” each player 50% of the way to the max games played. For a guy who is constantly missing a day here and there, in a weekly league, you won’t be able to sub in a replacement. Same for catchers; if they get benched a lot or they sit out a lot of games, you don’t get to sub in a backup. Also, it should top up the games played rather than plate appearances. Otherwise a guy who has a low PA projection because he’ll be batting 9th is getting bumped up to match the plate appearances of a guy who bats leadoff. This devalues leadoff hitters.
And another question: how does it take bench players into account? The value of bench players really depends on whether a league is daily or weekly, and whether or not they are acquired as part of a regular draft/auction or in a separate reserve draft.
Overall I’m really glad to see something like this returning. Lastplayerpicked was a godsend for quick dollar values. Thanks for the hard work! It’s bittersweet having a nice tool like this, and then realizing everybody you play against now also has that tool!
Calculator seems to work well when default settings select Steamer, but after setting up my scoring and trying to change to Fans, I get an error page…?
Can you link me to your scoring system? The FANS projections have considerably less players, so it would be helpful to see your exact league setup.
Just wanted to say thanks for this. As a free tool, this is far better than 99% of the tools I’ve seen that are pay features.
Fangraphs is seriously the best site going in regards to content they provide their users.
now i am being stupid, so excuse me.
what is the meaning of the following? mAVG, mRBI and so on? I guess i simply do not understand the results. If i have Denard Span @ 17, what do the PTS, aPOS, Dollars, and Adjusted costs mean?
Again, sorry for being clueless.
This is the greatest thing ever. I will nominate you for a Nobel Prize. I’m not sure which one, but rest assured, it will be one of them. (Probably the Peace prize, if Arafat and Obama can win one, I’m guessing they don’t screen candidates very well.) Also, I failed the first robot test. What does this mean?
Awesome tool. It would be really helpful if there was a way to account for supply and demand. For instance, in an AL only league if someone kept Felix, Sale, and Price for $1 each, you would expect to see someone like Kluber go up like $5-7, not 50 cents. I think the adjustment that calculates once you enter keepers should impact the elite players that remain far more disproportionately than some random RP who probably wouldn’t be impacted by these 3 SP being kept.
This is going to depend on how you approach the auction in general. If your goal is to get as much value for players as possible, then your described approach is probably not something you want to do.
With this calculator, you can get a sense for what players are “actually” worth as opposed to the economics of your individual league.
If you want to potentially overbid on scarcity, by all means, that’s something you can do, and in certain circumstances it might be the correct thing to do, but I don’t think that’s something we’re going to build into the calculator by default. Maybe as another experimental option down the road.
Way over values catchers.
It will depend on your league settings. If you’re in a 2 catcher league, the replacement level catcher is going to be pretty low, bringing up the value of catchers in general. Also keep in mind these are all calculated values and not necessarily market values.
If you want market values, go look at AAV’s. But if you play in a league where catchers are going for less than their actual value, it’s a market inefficiency you might be able to take advantage of.
David,
Thanks for this great work. Something fishy seems to be going on with Strikeouts for relievers. I think this needs an innings limit field to serve as the denominator, because otherwise relievers are showing up as negative in strikeouts….
Generally, all relievers are going to show up negative in strikeouts compared to starters. This is actually something I went back and forth on a little bit. Should relievers be baselined against relievers or “average pitchers”. In the end, it doesn’t really matter because it’s either corrected in the positional adjustment, or in the initial baseline.
The reason we are correcting things in the positional adjustment is it makes things a lot simpler for dual eligiblity pitchers and if you play in a league where there are not individual SP / RP slots and there are just Pitcher slots, then the numbers make a lot more sense.
Terrific tool, thank you thank you thank you. Having a bit of trouble getting to pitcher values that make sense for my league, but one thing that would help is a minimum innings option (my league ha a 900 minimum innings requirement). Otherwise this keeps suggesting too many relievers to be drafted than is realistic. I tried using the option to reduce the value of relievers by various amounts, but this just shoots up the value of the ace starters to nose-bleed levels in order to get positive values for the #3-4 starters that are sure to get drafted in my league. So please add a minimum innings option to the list of requests from your humble and grateful readership!
I’ll take a look at doing a minimum inning requirement. I’ll have to think about that a bit though, as I think correctly valuing pitchers up to a certain number of innings might be a little tricky.
One thing you might consider doing is if you play in a league that only has pitcher slots and not individual SP/RP slots, you could set it to say 4 SP slots and 1 RP slot. This might have a slightly different impact on decreasing RP value than the artificially decrease RP option.
Yes, as I was typing my comment, I was thinking I should just use 4 starters as a minimum, since it is essentially a practical minimum for my league. I just want the $1+ pitchers to represent enough total innings to be realistic, so I’ll play around with that. Agree that valuing pitchers in this way would be tricky.
This is essentially what I did with the old LPP site. Keep converting SP to RP slots until I had a good amount of IP per team. I think they could even use fractions as long as the total player pool was an integer (ie: 3.5 SP per team = 42 SP in a 12 team league).
Another method is including IP as a category, since we’re ultimately trying to give more value to players that provide more volume. That way, a swingman projected for 100+ IP doesn’t completely miss out on the value bump for SP’s, and it gives starters projected for 200+ IP more of a bump than SP’s projected for only 140 IP.
Just as a quick aside to all participating in the comments, the way we are calculating auction values is by no means, the only way to calculate them.
When you make an auction calculator you have to build in a bunch of assumptions and make a bunch of choices along the way, especially to accommodate a wide variety of leagues.
Most of the difficult decisions are made around baselines, standard deviations, positional adjustments, and replacement level. Exactly how you want to set those is up for debate and there is generally a lot of debate about how to do that. If you do it in a reasonably sensible way, you will most likely come out with somewhat similar results to other sensible choices.
In future versions, I would actually like to have a way for users to enter in the standard deviations of their individual league in each category. This is probably something for next season, but I think there’s a lot of “advanced” options that can be added and I’m definitely taking notes on this thread.
Are the values of SP and RP already scaled/discounted to reflect that SP cannot accumulate SV and to somewhat ignore RP W?
Also, Is there a way to change Streamer PA to the FAN projections while keeping streamer’s rates per PA? For a guy like Mookie, for example, I think the PA total is too low, but I think the rates are fine (or perhaps a tad bullish).
Ah, just saw the “Experimental” feature which wasnt showing up previously. I wonder how this weights against the value of dead roster weight — a discount value of a lost spot on a limited roster due to platooning, etc.
If an SP is projected for 0 saves, he will not get credit for any saves. If a reliever is projected for 3 or 4 wins, he will get credit for those, as he should. There’s no reason to ignore those wins. In fact, relievers and starters get wins at almost the exact same rate (about 15 IP per win) and in leagues that have IP maximums, relievers are very valuable for their K and W totals since all categories are effectively rate stats (K/9, W/9, etc)
Sorry — one other notable item. When you change the projections from Batter to Pitcher or to some other selection, all selected criteria “reset” — might I suggest “remembering” those settings as you toggle between the positions/type of player?
All the league settings / keeper settings should persist across tabs. Can you link me to a page where you if you click on pitchers / batters your league settings reset?
I wonder if this might be because the tool is so popular/data-use heavy that it was just crashing the browser? I’ll try it on my home network to see if theres any difference and report back.
I could have sworn I posted this yesterday, but I would love the ability to separate OF into LF/CF/RF.
Yes, this!
fantasy n00b question… Do these auction values correlate to a snake draft order? I.E. Will it work to place the highest auction $$ value as my top draft order guy, second highest dollar amount as player 2, etc?
Dave – the Last Player Picked (and the SGP calculation) was some flawed logic. Players in the bottom of the universe of chosen players are not $1, they’re < $1 – typically negative value – BUT someone must bid a buck on those guys (hence a forced overbid).
If you have X amount of dollars to divide into Y players, then you are going to get the average dollar value of what a player (hitter/pitcher separately) is worth. That is one data point. A typical league will bid $1 on, say, 30 players, so you have to draw that line between the average player and the "typical" $1 player, but it shouldn't be the worst of the $1 players, it should be near the top of those 30 players. That should be your second data point and all dollar values of all other players should be linearly interpolated to get to a dollar value.
To make the point another way: Suppose there is a player who is projected to get 1 AB for the entire season but he MUST be bid on (forget the reason why for a moment). If that was the case, would he really be worth $1? Definitely not.
This is the best, most customizable auc calc I’ve seen! Thank you! Would love to be able to export with real projections instead of the m(proj) that are there. Can you make that a setting? Thank you!!!
Is there a maximum number of keepers you can enter? I am in a 14 team dynasty league with 40 man rosters, and I have been entering keepers for hours, down to the last 2 teams and I can’t get any names to show in the player finder anymore. Any explanation for what might be causing this?
We are looking into this issue. I’ve been able to replicate the error, but don’t have a fix at this time.
Excellent, Thank you!
PS – This resource is amazing and saving me a ton of Excel spreadsheet time. Keep up the great work and again, THANK YOU.
I was expecting the positional adjustment to be negative for deep positions like OF and 1B. But it appears that the aPos column is positive for all players. Am I misunderstanding something?
It’s really going to depend on how many players you have starting at each position and the league settings, but in general, the baselines for the standard scores are most likely going to put all the replacement level players at each position at a negative standard score. This will make all the positional adjustments positive, but that’s nothing to worry about.
Any way for us to see some of the underlying formulae?
I’m having a hard time trusting the slash stats values. To pick out 3 names and their SLG, all of them projected for negative value, it says Altuve (410 SLG in 618 ABs) is worth slightly more than Ryan Howard (400 in 402 ABs) who is worth slightly more than Wil Myers (412 in 496 AB)
If 412 is a negative value, then it might be that Myers extra ~100 ABs makes him a bigger drain on SLG than Howard. But in that case, Altuve should be an even bigger drag, worth less than both of them.
If, for some reason, the ‘zero’ is actually set somewhere around $-3, so 410 SLG is a positive, then it makes sense that Altuve is just ahead of Myers, the extra 120 ABs of positive value outweigh the fact that Myers’ projection is 2 points better. But in that case, Howard should be well below both, as each AB is less positive value, and there are less of them.
It’s great having the AV calculator there, definitely easier than trying to export all the steamer/fan projection myself. But unfortunately I can’t trust them yet, because they’re outputting numbers that look clearly mistaken to me, I’m pretty sure there’s something up with how it is calculating rate stats with different volumes.
If you just take SLG as the only stat with the default other settings, the formula is going to be:
(((1B+2B*2+3B*3+HR*4) – AB * 0.4615) – -0.00507738) / 11.01100723
.4615 is the average SLG of the top x players, -.00507738 is the average xSLG. and 11.011 is the STDDEV of the xSLG.
If you just run their xSLG you get:
Howard: -24.5
Myers: -27.9
Altuve: -30.2
Altuve is actually going to hurt you the most because he has the most AB and all of them are below average in SLG.
But, once you do the positional adjustments the order looks like this:
Howard: -$10.40
Altuve: -$10.60
Myers: -$10.90
This is of course going to vary greatly depending on your league settings.
I hope that helps clarify things. If you still feel there’s a problem, let me know.
Thanks.
I get the same xSLG as you do when calculating by hand, but with default settings and just SLG added, the mSLG order generated by the calculator is Altuve > Howard > Myers.
If the average SLG is 0.4615, so those with projected SLG of .4615 have xSLG of 0, why is it that the players with mSLG of roughly 0 all have SLG of ~0.430-0.435?
There’s a positional adjustment factor being applied within each stat, as well as the extra positional adjustment being applied at the end? I think thatmight be what’s causing the results I see as problematic. (and just tested by keeping things the same except the league having 5 2B, and Altuve’s mSLG jumped well ahead of Myers’ & Howard’s, so I assume that is what’s happening.)
Is that positional adjustment only applied to the rate stats? A quick look suggests it’s not applied to counting stats, everyone projected for 83 RBI gets mRBI of $3.1, regardless of whether they’re 1B, 2B or 3B.
Assuming the extra positional adjustments that seem to be applied to rate but not counting stats are deliberate, is there any way to avoid that, to simply generate values for each stat irrespective of position? I can’t do it by setting all the position requirements to 0 except with 15 Utilities, as it’s capped at 10. (And surprisingly, making all batters equal still gives different values for mPOS.) I’d much prefer to be able to get values that ignore position, and then add in overall positional adjustment afterwards.
How easy would it be to add a row that showed what the average being used for each rate stat was, so it’s easier to see what SLG, AVG, etc will give an xSLG, xAVG of 0?
All the players right around 0 mSLG will really have a SLG around .461. It’s possible you are seeing players right around 0 have a lower SLG because they have limited plate appearances? An example of a particular player would help.
There is only a positional adjustment being applied to each player at the end. It is not applied for each stat. All players are judged against a .461 baseline in the above example. The aPOS column is the positional adjustment that is applied at the end.
Using default positions (14 batters, 1 2B, 1 1B, 5 OF), just adding SLG:
Everyone with an mSLG of 0.0 and 100+ ABs is projected for SLG of either 0.432 or 0.431. http://www.fangraphs.com/auctiontool.aspx?type=bat&proj=steamer&pos=2%2c1%2c1%2c1%2c5%2c1%2c1%2c1%2c0%2c1%2c6%2c3%2c0%2c10&dollars=260&teams=12&mp=20&msp=5&mrp=5&mb=1&split=&points=c|0%2c1%2c2%2c3%2c4%2c6|0%2c1%2c2%2c3%2c4&lg=MLB&rep=0&drp=0&players=&sort=9%2cd
The players are Melky Cabrera, d’Arnaud, Carlos Santana, A. Gordon, Calhoun, Lambo, Pence, Frazier.
Looks like the issue with Altuve’s mSLG being higher than Myers’ was coming from rounding errors, as it has now disappeared.
This might be an issue caused by having empty spots/breaking the calculator with an extreme setting, but making all batters eligible as Utility only has mPOS thinking that 2B is worth almost $3 more than SS for the same stats, and David Ortiz and any other DH-only players are worth ~$13 less than everyone else: http://www.fangraphs.com/auctiontool.aspx?type=bat&proj=steamer&pos=0,0,0,0,0,0,0,0,0,8,6,3,0,10&dollars=260&teams=12&mp=20&msp=5&mrp=5&mb=1&split=&points=c|0,1,2,3,4,6|0,1,2,3,4&lg=MLB&rep=0&drp=0&players=
In the default positions, just adding SLG for the first link you provide, the average SLG for the top x players is .430, so that all seems like it checks out.
That is a good catch on that very strange league configuration. This should be corrected now.
Sorry to be a pain. For the first bit, I’m wondering if it was a simple typo, and the actual average for SLG is 0.4315? That’d explain why ~0 mSLG happens with SLG of around 430.
Annnd, there’s some weirdness with Steamer. Altuve is projected for 136 1B, 37 2B, 3 3B, 9 HR in 618 AB. That’s 255/618 = 0.413 SLG. But his player page also says steamer has him down for 0.410 SLG. It says Myers is down for 78 1B, 24 2B, 1 3B, 18 HR in 496 AB. That’s 201/496 = 0.405 SLG, not the 0.412 it’s actually listed as. Probably a separate discussion, but any ideas about why there’s such a discrepancy between steamer’s projected SLG number, and the SLG generated by taking steamer’s projected TB/AB?
Thanks for pointing this out. Steamer actually calculates all raw stats to the decimal place, which we round for display purposes. But we calculate all the rate stats using the decimal place. This is why if you try to calculate SLG or another stat using the stats on the player page, they will be slightly off.
I had forgotten about this and was pulling the rounded raw stats. In practice this might change the auction values by $1 maybe, but it’s better to do things the right way, and it will avoid confusion in the future. The auction calculator is now pulling the non-rounded stats.
Is there any chance that a feature can be added to actually use 2014 stats instead of just different projections? Seems like this wouldn’t be too difficult. Projecting performance is only half of the game in fantasy. The other half is figuring out how to properly value a given performance. It would be nice to be able to see this half on its own.
There seems to be an issue with how pitchers are evaluated. I am in an old school 12 team AL only 4X4 league. When these settings are input even with a minimum of 6 starting pitchers, there are multiple pitchers projected for very low innings, for example Yohan Pino and his 10 projected innings are the 39th best starting pitcher, being valued at almost $5.
This should be corrected now. Yohan Pino was causing some issues with position eligibility because he was a starter all last year and was being consider SP eligible, even though his stats were projected as relief pitcher stats, which basically made him a neutral, if not slightly above neutral ERA and WHIP candidate.
What is the category ‘aPOS’?
It is the positional adjustment to account for position scarcity.
It’s been said above, but it would be awesome if you could add in LF/CF/RF as an option instead of just OF’s. All 3 of my leagues use the split, and more and more leagues I’m seeing use this as well. I appreciate the effort put into this, and it looks like a fantastic tool. But it just wouldn’t make sense for me to use this, unless I wanted to manually adjust all the OF’s based on their scarcity. I could do this, but currently I’d rather use tools from other sites that allow for this distinction.
Thanks David.
I have a tool that calculates value in a very similar way and the only real differences I’m seeing in value relate to AVG (perhaps other rate stats too). Out of curiosity, do you adjust the mAVG for the number of plate appearances (reflecting that a .240 AVG from a guy with 400 PA’s doesn’t hurt as bad as if it is from 600 PA’s. Given that you show PA’s, I’m thinking you do. Perhaps I’m “over” adjusting for it.
I’m in a 15 team league with a $260 budget and 30 players per team. The total amount spent should be $3,900. However, when I use the auction calculator, there’s no combination of hitters and pitchers that totals $3,900. It does get within $150 or so. Just an FYI.
David, first of all thanks for creating this tool. I’ve been working on something similar for my league for the past couple years, but yours looks to be simpler and more powerful. Awesome stuff.
I do have a couple of questions:
1. If I am entering keepers, does the calculator deduct the cost of those players from the total draft cash pool? For the sake of a simple example, let’s say I play in a 4-team league and each team has a $100 draft/keeper budget. Each team keeps one player apiece, and all these players are $5. So that leaves each team with an adjusted $95 “cash on hand” to spend in the draft. Does the calculator account for this automatically, or is it still going to assume $400 ($100 x 4 teams) is going to be spent in the draft rather than $380 ($95 x 4 teams)? And for purposes of calculating player value, which approach is more accurate anyway?
2. Is there any way to add custom weights to each statistical category? To pick one example, in my league (16 teams) we weight the HR category 2x the RBI category. So you get 32 points for finishing first in HR, 30 points for second, 28 points for third, and so on. You get 16 points for finish first in RBI, 15 points for second, 14 points for third, and so on. I guess I could probably fudge this manually in Excel with the calculator’s exported values, but it would be a nice feature to add for future versions of the calculator.
I don’t know if this is an issue but something that seems weird. I’m playing in a league using OBP and TB, calculated my own auction values. I used the auction calculator to make sure I wasn’t crazy and for the most part the range of values line up with just a few small differences based on projection.
The major exception is TB. The fangraphs auction calculator puts a ton of $ value into TB. While my $ values for TB have a similar range to $ values for other categories, about $14 for the best in TB Mike Trout, the auction calculator range goes up to $22.9 for Mike Trout. I’m trying to think of the reason for this.
I appreciate the QS category that got added, but I noticed that a number of pitchers don’t show up if using QS instead of W. Maybe that’s because they won’t have many QS, but it seems unlikely that Trevor Bauer’s lack of efficiency would keep him out of the top 500 available pitchers. I also noticed Wade Miley’s absence and a couple others.
Thanks for this!
I was playing around with your handy auction calculator using my NL only league settings and noticed an anomaly. Looking down the list of most valuable pitchers, sticking out like a sore thumb is Yusmeiro Petit at position #10 – right between Jose Fernandez and Zack Greinke.
While I don’t doubt that he is a fine pitcher, I know he is not going to go for $28 in my $300 salary cap league. I would guess he will go for $5 or less. Why is that?
I have to assume that most people avoid pitchers who don’t accumulate saves OR wins. With only 6 pitcher slots and a minimum inning requirement in my league, nobody wants to throw somebody in their lineup who will essentially take up a closer spot while accumulating zero saves.
So I guess I’m wondering: is the reason for this discrepancy completely based in innings requirements that are not accounted for in the valuation system? Or is there something else going on here? What is a good way to account for innings requirements in a calculator like this?
Immediately after writing this, I realized that there are lots of pitchers listed as SPs and receiving the SP positional adjustment when they should probably be listed as RPs. Seems odd.
Could we get quality starts added for points leagues?
ZIPS is still not added to the auction calculator. is that going to happen soon? thanks
Unless I’m doing something entirely wrong, there appears to be a big huge massive problem with the calculator, at least with respect to the starting pitcher population. (i haven’t checked the hitters… and the relievers seem OK)
For example, if check the starters tab, and sort by team, I don’t find a single Rockie, and there’s only three Royals (Ventura, Duffy, Vargas). Now, fine, I probably shouldn’t be going into my draft looking to buy any Rockie starters, but there could be a conceivable reason for someone to look for a valuation on Jeremy Guthrie or Chris Young or Edinson Volquez.
Is it somehow limiting the number of players in the pool based on a lower level cutoff value for some of the stat categories?
The biggest issues seem to be if I include home runs allowed in the mix, as this strips away about 30 pitchers from the pool….
Indeed, if I set up the calculator with the 4×4 Otto Classic settings, with ERA, WHIP, SO, and Hr (because HR/9 isn’t available), I get a pool of 137 starters, and only one of them is a Royal (no yordano ventura, for instance, and only two RedSox starters, neither with more than 40 projectd IP).
I’d love to be able to use this calculator, but unless there’s some suitable and reasonable explanation for the missing pitchers, I can’t trust it.
Can you add the playerid field to the export?
I love the calculator. Thanks!
Do the projections in the calculator update to rest of season projections as the season moves along or could that be an option. That was a nice feature of the Hardball Times calculator because you could see how players’ values rose and fell as the season wore on. Thanks!
Came here to say the same thing. The ROS projections are really useful when determining trade values.