Projected Win Totals, Graphed
This post is a bit of an experiment. During the day, while working on various writing topics and filtering through different ideas, I’ll often come up with something that isn’t quite worthy of a full 1,000 word post, but is interesting enough to share on its own. A lot of times, these things just end up on Twitter, but sometimes, they just don’t go anywhere.
Instead of just leaving these in the back room, I’m going to start putting them up just as stand-alone, low-commentary posts and see if there’s enough interest in the data points as conversation generators to continue posting them. If it turns out that you guys don’t like them, I won’t keep doing them, so critical feedback is certainly welcome, but perhaps there’s room on FanGraphs for posts that aren’t quite fully fleshed ideas, but instead just interesting statistical nuggets. We’ll see, I guess.
If you go to the site’s Playoff Odds page, you’ll see some pretty staggering numbers for the teams at the top. Our model currently forecasts the Tigers to have a 94% chance of reaching the postseason, for instance, both Bay Area teams are strong favorites to reach at least the Wild Card game as well. These numbers are surprisingly high given that we’re still in May, and there’s fourth months of baseball left to play. A lot can happen in four months.
But to illustrate why those numbers are so high, it helps to take a look at the projected final records in graph form, because there are some huge gaps between the teams at the top and the teams in the middle or bottom.
Here’s the American League first.
And now the National League.
The favorites in both league have very high playoff odds figures, but for different reasons.
In the AL, it’s because there is a huge gap in expected win totals between the top four and the next eight, with an overstuffed middle class fighting for what looks like one wild card spot, with none of them expected to be all that close to catching any of the current division leaders.
In the NL, the story is essentially the opposite, as there is basically no middle class to speak of. There are good teams and there are bad teams, and not much in between. While the Pirates aren’t completely dead in the water yet, it looks like the Brewers and Rockies are the only super serious threats to overtaking the established Big Five.
You can see the differences in league parity when the two leagues are plotted together.
Look at how weak the back half of the NL is in comparison to the second half of the AL; there are a half dozen NL teams that already have no real chance at reaching the playoffs. In the AL, pretty much everyone besides the Astros, Twins, and White Sox is considered a possible playoff team, but the huge blob in the middle means there’s no real serious outside threat to the top teams.
Injuries, trades, and just the unpredictable nature of the game will likely make it so that something unpredictable happens between now and the end of the season, but I would not be surprised at all if the nine teams that look to be strong favorites to reach the postseason all end up reaching that goal.
Dave is the Managing Editor of FanGraphs.
The fact that the Jays are in the “strong favourites” part of this discussion both incredibly excites and absolutely terrifies me, and presumably all Jays fans. I think I speak for more than one when I say that we are all holding our breath and not trying to appear too excited, but are going absolutely bat-crap crazy freaking out on the inside.
my sentiments exactly.
As a biased Yankees fan, my feeling when I look at the Jays is that their pitching isn’t good enough for it to last. Of course, I also think the Yankee’s pitching will magically come together, and Beltran will be back any day now and will finish the season as an MVP candidate. …so, you are probably fine.
The general perception of the jays’ rotation seems a bit strange to me.
I get that they blew up last year but thesecare all still legit mlb arms with legit track records…..and that’s true even if we ignore all the good years dickey/buehrle/happ had before becoming blue jays.
Here’s how the jays’ SP have done as SP even just with the Jays:
Buehrle 43gs, 6.3ip/gs, 3.66era, 3.86fip, 4.08xfip
Dickey 45gs, 6.5ip/gs, 4.15era, 4.43fip, 4.27xfip
Hutch 22gs, 5.7ip/gs, 4.22era, 4.17fip, 3.96xfip
Happ 29gs, 5.2ip/gs, 4.34era, 4.02fip, 4.46xfip
Morrow 93gs, 5.7ip/gs, 4.40era, 3.71fip, 3.76xfip
And then the career AAA numbers of the depth arms:
Stroman (23): 35.7ip, 11.4k/9, 2.3bb/9, 3.03era, 2.09fip
Nolin (24): 57.7ip, 7.3k/9, 4.1bb/9, 2.34era, 3.56fip
Hendriks (25): 302.7ip, 6.2k/9, 1.5bb/9, 3.27era, 3.11fip
Maybe not awesome, but good enough that the predictions of impending implosion seem a bit overblown.
My perception of them is largely because when I watch them, they are facing the Yankees, and the Yankees have had a lot of success against them recently. Also, they lack clear front of the rotation talent. I agree that they are probably good enough to hang around though.
Fair enough, though buehrle’s doing a fair impression of a top of the rotation guy, while dickey won a cy a year and a half ago.
Buehrle has been spectacular. I hope he keeps it up (even if it will hurt the Yankees). I love his approach to pitching. Sadly, Dickey seems nothing like the guy who won a Cy a few seasons ago, because the knuckleball is and will forever be the greatest pitch.
Don’t look now but dickey is 4-1 in his last 6 starts with a 2.52era, .574oops, and 6.6ip/gs.
Ahh, good for him. ….but rooting for both Dickey and Buehrle is too much for this Yankees fan. One of them will have to fall off a cliff…
No one said “strong favourites.”
Damned canuckians.
Huh?
“…but I would not be surprised at all if the nine teams that look to be strong favorites to reach the postseason all end up reaching that goal.”
To be fair, Dave spelled it properly and didn’t use the poncy Brit spelling.
Bah. I completely missed on first read that your comment was about the extra “u”, and not the actual content of said comment.
Pretty silly, eh?
You guys need Hutchison to not be the guy he was yesterday.
Blue Jays will make the playoffs if they stay healthy. That lineup is just too good. Also, they will likely add another SP and possibly a 2b at deadline.
I think I could triple slash a .001/.0015/.005 for them at 2B, and commit a maximum of 25 errors per nine innings.
Which puts me in line for a September call-up! I think I’m like 6th on the depth chart currently.
LOL when Happ walked Crisp on 4 pitches Sunday, I thought they were done. My faith in them is a kilometer wide and a millimeter deep.
“If it turns out that you guys don’t like them, I won’t keep doing them, so critical feedback is certainly welcome, but perhaps there’s room on FanGraphs for posts that aren’t quite fully fleshed ideas, but instead just interesting statistical nuggets.”
If you did the legwork, I do not see a reason to hide the results. Maybe these eventually become FG+ content. Maybe these posts are titled with a notation before or after the article name to allow uninterested viewers to save time. Sharing new ideas is generally a strength of Fangraphs, and I see this type of article as development of that strength. Sharing abstract and theoretical ideas about development or creation of statistics would allow interested parties to further research that might otherwise go dormant.
yeah I like the idea of calling these like “Nuggets” or something so that people know they’ll be brief. i’m a fan of this though, more quality content is pretty much always good, even if it’s smallish.
Can I just offer a couple of chart suggestions?
First, bar charts should always have a y-axis that starts at 0. The way we process bar charts visually is to compare area. Using that first glance your top graph makes it looks like Det expected wins are roughly 10x Hou.
Second, final second graph might work better as a line graph. I think it would be easier to compare the values to each other if they were expressed as two lines rather than a series of paired columns (I know lines are usually for looking at trend, but in this case we’ve already established the bar algorithm, so less chance for confusion).
It’s because of the whole “everyone wins 50 and loses 50” (or whatever the number is). If the y axis started at 0 it would be pretty hard to differentiate between all the teams.
There are only two teams since 1962 who lost more than 50 games. Furthermore, replacement level is sort of defined as about 48 wins so it would seem that we assume that each team is “entitled” to those wins simply by showing up with 25 guys who are worth the league minimum (sorry, 1962 Mets and 2003 Tigers).
So starting the graph at 50 would capture the scale of the differences between the teams far better than starting at zero.
I second Andrew, respectfully disagree with Goofball. Keep it as-is.
I am strongly in favor of posts like this, whether by Dave Cameron or any other FanGraphs writer.
I don’t see why they need any special notation. I almost always decide whether or not to read a post by its headline.
i am all for more content, especially when it is so concisely and efficiently conveyed. sometimes a short post is better than a long one.
I always find it interesting to look at gaps between the actual/historical and the expected/future. So, the interesting teams to me in that sense are those with the biggest differences, in this case the absolute value of the difference between win% YTD and projected win% for the rest of the season:
RedSox 11.6% better
Giants -10.4% worse
Brewers -9.1%
D’Backs 9% the projections for the Brewers and DBacks for the remainder of the season are ~the same.
Rays 8%
Cubs 8%
As -7%
Interestingly (to me) the Braves are expected to score waaay fewer runs, but are also expected to allow waaay fewer runs (second biggest shift in expectations in each category), with the net result a wash.
“Interestingly (to me) the Braves are expected to score waaay fewer runs, but are also expected to allow waaay fewer runs (second biggest shift in expectations in each category), with the net result a wash.”
I think you have that backwards if you’re intending to compare results to this point to projected results going forward. The Braves are expected to both score and allow more runs per game than they have to this point.
you are correct.
A’s +100 runs second best Angels +39 Tigers +19
A’s 2nd in HR first in BB first in HR allowed 4th in BB allowed
2nd in runs scored first in runs allowed
Deep club
So why aren’t they out projecting the lowly +19 Tigers when they have the best record since the start of 2012 in MLB, because your projection system seems to hate A’s baseball and always underestimate them, and this is not new…
Haha this is one of the first times I’ve seen a team’s fan* belong… to the As. And accuse the system of bias against the As. Usually it’s the Braves, Giants, Yankees, and to a lesser extent, BoSox.
Say it ain’t so, Billy.
* fan is short for fanatic, so it makes sense that people feel biased against, I suppose.
My understanding of the etymology is that ‘fan’ is short for ‘fancier’, and was first used in the 19th century.
The attribution to ‘fanatic’ is retrospective.
With Apologies for the Pedantry,
A Linguistic Fanatic
You just may be right!
Warmest regards,
A linguistic fancier
Check it out, this guy is very well versed for a DJ.
Detroit is a 93-win team? Ahh, I don’t think so.
You’re right. The Tigers will probably not win exactly 93 games.
However, their chances of winning more or fewer are about even.
That’s why I asked “Detroit is a 93-win team?” and not “Detroit will win exactly 93 games?”
No way is Detroit even money to win 93 or more games. With a very shaky bullpen, a couple of big holes in the lineup, a couple of bad defensive players (especially Torii Hunter), and a very bad version of Verlander with no end in sight, they’d have to go 65-50 from here on out to win 93 games. Not saying it’s not possible – of course it is – but I’ll take the under.
Going 65-50 is only (only?) (only!) a .565 win percentage, which seems doable, particularly when you get to load up against the likes of MIN and CWS.
Again, referring to what is likely, as opposed to what is doable.
I was too.
Professional oddsmaker here (with no personal bias for or against Detroit).
Sure they are.
I don’t care if it’s petty to come back to this old thread, but I’d like to point out that the Tigers need to go 30-17 from now on in order to win 93 games. It’s mainly due to their terrible bullpen, the useless offensive players they throw into the bottom of the lineup, and bad defense by Torii Hunter. LIKE I SAID.
Professional oddsmaker. Puh-lease.
“Interesting statistical nuggets” are the #1 reason I read about baseball and enjoy talking about it. Thus, it goes to follow that these posts are awesome.
Except, I don’t like Jayson Stark-style statistical nuggets, he ruins them for me because he follows every single one with some sort of blathering nonsense like, “This is why baseball is the greatest game!” or something that is completely unrelated to what he was actually saying. Jayson, get it in your head, we’re not reading your column to decide if we like baseball or not! We don’t need to know that! STOP IT!
Seeing as that is basically Jayson Stark’s core attribute, I would suggest you stop reading his articles.
Please keep the articles coming!
No love for my Twinkies, last next to the ‘Stros. They are going to show everyone and get that second WC, I’m feeling lucky. Yankees will eventually realize they are all old and fall apart, Angels pitching has serious regression written all over it. Seattle, come on man. The rest of the central is anyone game after Detroit.
You heard it first right here.
“You heard it first right here.”
And last, probably.
It seems like there will be many teams looking to add one or two pieces for the stretch run…
Could be a fun trading deadline.
I didn’t even read this post yet, but after the first two paras, I figured I’d throw my vote in for “more please.” Sounds like a solid idea to me.
This is nice content, Dave. I would like to see more posts like this, maybe an “FG Shorts: (Title)” series. If this becomes a regular thing, it would be nice to easily distinguish shorter posts.