Your Stance On the Team Projections
Out of all the polling projects I run, this one’s always my favorite. No sense in beating around the bush. Here are our current, schedule-adjusted projected 2018 standings. These are based on the Steamer projection system, the ZiPS projection system, and manually-adjusted team depth charts. Now that the Orioles have signed Alex Cobb, there are hardly more big changes to make between now and opening day. So, what do you actually think of the projections you’re seeing?
For convenience, here are the league landscapes, in case you didn’t feel like clicking the link.
The top looks like the top you’d expect. The bottom looks like the bottom you’d expect. We’ve been writing about the various tiers for months. But, even if you might not realize it, you’re experts. You know a lot about particular baseball teams, information the projection systems might not be aware of. So you might consider certain team projections too optimistic, or you might consider certain team projections too pessimistic. This is your collective opportunity to make yourselves heard. Last year, the community thought the projections were too low on the Rockies, Brewers, and Royals. All three teams won more games than was projected. Meanwhile, the community thought the projections were too high on the A’s, Angels, and Marlins. All three teams won fewer games than was projected. You all can provide valuable input, and so I love when this project gets to the analysis part.
The analysis part is coming, probably early next week. Following, you’ll find 30 polls for 30 teams. They should be simple to understand, especially if you’ve done this before. Vote based on research, or vote based on gut. I don’t care. Just vote. All I ask is that you vote based on the information we know today. Vote based on the rosters and depth teams have, and don’t vote based on the assumption that a team will make midseason additions or subtractions. That stuff is effectively un-projectable. Everything now being said, I leave it to you. Thank you all in advance for your participation.
To proceed directly to a specific team projection poll, click on the team’s name below.
Angels, Astros, Athletics, Blue Jays, Braves, Brewers, Cardinals, Cubs, Diamondbacks, Dodgers, Giants, Indians, Mariners, Marlins, Mets, Nationals, Orioles, Padres, Phillies, Pirates, Rangers, Rays, Red Sox, Reds, Rockies, Royals, Tigers, Twins, White Sox, Yankees
=====
Angels
Astros
Athletics
Blue Jays
Braves
Brewers
Cardinals
Cubs
Diamondbacks
Dodgers
Giants
Indians
Mariners
Marlins
Mets
Nationals
Orioles
Padres
Phillies
Pirates
Rangers
Rays
Red Sox
Reds
Rockies
Royals
Tigers
Twins
White Sox
Yankees
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
There is zero chance SF finishes better than AZ. 9/10 times AZ should be projected higher than SF.
Okay, 5% chance. Sorry.
Your math is still off. You said there’s zero chance, and then you said 9/10, and then 5% chance. Which one is it?!
There’s a 0% chance that 9 times out of 10 Arizona should be projected to finish ahead of the Giants 5% of the time.
Is “should” a normative statement or a statement of probability?
I’d take the Diamondbacks over the Giants, but no way I’d give someone 19:1 or 10:1 on that bet. In fact, I’d gladly wager $$$$ on the Giants at that price.
60% of the time that math works every time
I’m trying to avoid doing electioneering in the comments because it might bias the results but I also found the early voting on the Giants suprising.
Yankees are way low. They are winning 163 games and hitting 500 homers this year confirmed.
Ouch – clean sweep of the season but they still end up getting bounced in the Wild Card game. Tough way to end a magical season.
If they swept the season they’d have to win the wild card game to get that 163rd win.
But they can’t play in the wild card game, because going 162-0 means they won the division (since they play the other teams in their division).
So that means they go 162-0, and lose the ALDS 3-1 (assuming your “clean sweep of the season” comment is correct).
I really want to have more faith in the Rockies than 79 wins, but I just do not trust the offense.
Pretty sure I would never say that.
There should have been an option of “all of the above” for the Mets.
Would love to see a separate poll on “How confident are you about that?” for each team.
Something like that will probably be revealed by the distribution in votes for each club.
That seems unlikely to me. I abstained from a bunch of polls where I felt I wasn’t confident enough, and voted “good” in cases where I have both very high and very low confidence.
I made a comment that ties into this to a large extent – wondering what effect it would have on the voting distribution if people were required to bet a dollar on each projection.
You have to do some pretty detailed digging about the playing time assumptions to make an informed guess about the Mets.
I find it interesting that there is only one team (STL) with a playoff probability between 43% and 82%. Last year there were 6 such teams. In 2016 there were 8. The good teams are getting better (or getting more favorable projections). Vice versa with the bad teams.
It’s the “superteams and tanking” era
I’m hoping that teams choosing not to be competitive will pay dearly for those decisions. Until now, they haven’t paid for that choice. But it’s becoming a bigger problem that so few teams are actually in it from the start of the season.
Bad news, friend: they’re actually going to be rewarded
The thing, especially, to consider for the teams in the high-70s is what happens at the trade deadline when they decide that they’re not WC contenders. Parts will be moved and rosters at EOS will not be as competitive as they look now.
The flip side of that are the real contenders in weaker divisions. The Nats and Indians, IMHO, will be adding at the deadline while their division mates will largely be selling off. That could be worth a 1 or 2 win bump in the positive direction.
Yeah, that’s what I see with Detroit. They are already either The or One of The worst teams in baseball, and if Fulmer reestablishes some value, he’s gone, plus Iglesias, a couple relievers, maybe Castellanos, and any rental guys (like Fiers & Liriano).
This is a team that is going to be really, really bad after the trade deadline.
There isn’t much left to deal there. And the Tigers are one team that I expect to surprise some people. They are going to score a lot of runs if there middle of the order holds up – Miggy, Castellanos, VMart. I guess Nick could get dealt away but the market for that sort of bat hasn’t been strong in recent years. I could see them getting to 80 wins and hanging on the periphery of the WC for a while.
An interesting variation on this would be if you had to bet some nominal amount like a dollar on each selection you make. The people who get the most right win some amount of the total money, and the rest goes to a charity.
I wonder how much this would affect people’s predictions. Although I have to say that the consensus does seem pretty accurate to me in most cases.
Jeff, can you do a post about what the assumptions made by the projected standings? I know you mentioned in the podcast that they do not take strength of schedule into account, and had a post about the fact that on average they are off by 6-7 wins, but I don’t know what other oversights there may be. Injuries? Farm depth (and how is this quantified)?
Not quite. The projected standings (https://www.fangraphs.com/depthcharts.aspx?position=Standings) don’t consider schedule. But the numbers Jeff used are from the playoff odds page (https://www.fangraphs.com/standings/playoff-odds) which comes from playing the actual schedule 10,000 times or so.
6-7 wins is the variance of 162 events. If you flip a coin 162 times, you’d expect/project 81 heads. But you’d get exactly 81 heads ~6% of the time. So even if we had the end of season raw stats for every player, we couldn’t get projected win totals right. Sequencing and luck matter a lot.
The playing time models generally use average injury chance, relative to position and player history. This is why you’ll often see a guy who plays every day only projected for 90-95% of the playing time at a position. It’s not a projection that he’ll miss a few games, but that of all across the board SOME of the every day players will get hurt (see Trout, 2017) and that risk is averaged.
Having studied projections, they get depth wrong. You’d rather have 6 league average starters than 4 to fill the last 3 spots in your rotation. But projections wind up the same because no matter how you allocate playing time, it’s a league average performance. But that’s a minor error and only a few teams have players on their bench and AAA who aren’t replacement level.
Thanks for the informative response Adam! I didn’t know the playoff odds and projected standings were calculated differently.
I thought there may be some over/under season wins opportunity here. The best consensus is Tor under 85 and Mil over 79. Not surprisingly, the experts already reflect Tor 81.5 and Mil 83.5.