Baseball Prospectus recently came out with a prediction for the Mets record in 2015. They predicted that the Mets would win 82 games and lose 80. The question of course is whether this is possible or not. Potentially, this is actually a minimalist prediction.
We’ll start with the concept of Pythagorean Won-Loss. This is one of the most basic and interesting Sabermetric concepts out there. Without getting into too many crazy details and formula’s, it is basically a way of determining what a team’s win loss record should have been based upon the amount of runs they scored and allowed. Again, that is simplifying how the determination is made, but runs scored and allowed are the heart of the Pythagorean Won-Loss concept. Since 2011, the Mets have either perfectly matched the win totals they were supposed to obtain in a season or been a game or so off. In 2014, the Mets won 79 games, but Pythagorean W-L stated that they should have won 82 and lost 80. Sound familiar?
To get a better grip on this concept, let’s look at how run differential has related to National League teams over the past five years. Since 2010, 37 National League teams have had positive run differentials. Of those teams, only four have had positive run differentials and losing records: The Mets twice, last year and in 2010, along with the Marlins in 2010 and the Reds in 2011. That means that, approximately 90% of the time, if you score more runs than you allow, you win at least 81 games. Occasionally there are outliers in this as well, as over the past five years, three teams, the 2011 Giants, the 2013 Diamondbacks and the 2014 Brewers, allowed more runs than they scored and still finished with at least 81 wins, but you get the point. If you score and don’t allow the other team to score, you win. That’s the basis of Pythagorean W-L and of professional baseball in general.
Getting more deeply into the numbers, the obvious becomes, well, more obvious; as a teams’ run differential expands, the teams winning percentage generally also increases. Of the 37 teams that had positive run differentials since 2010, three had positive numbers less than ten. Two were the Marlins and Mets teams of 2010 that had losing records. The other was the Phillies club from 2012 that won exactly 81 games. Most teams that have tremendous success score at least 50 more runs than they allow. Since 2010, four of the 37 teams have scored anywhere between 50 and 60 more runs than the opposing team. Those four teams averaged 88 wins, with the high point being the Pirates of 2013, winning 94 games while scoring 57 more runs than the opposition and the low point being the Rockies of 2010, who won 83 games while scoring 53 more runs than their opponent. When the run differential hits the 100 mark, success is guaranteed. 13 of the 37 teams have achieved this feat, collectively averaging 94 wins with 11 of those teams making the playoffs.
So we come back to the 82 win prediction. Baseball Prospectus is basically stating that the Mets will play exactly as well as they did last year, except now they will actually play to that potential and win 82 games. The flaw with that concept is that it’s much more likely that the Mets eclipse the 82 win mark for a multitude of reasons.
To start, Matt Harvey is returning and any replacement starters for injuries will be made by the likes of Rafael Montero, Noah Syndergaard or Matthew Bowman instead of the various retreads the Mets have put out there in recent seasons. Let’s say that Harvey directly replaces Dillon Gee’s 22 starts from last year. Gee posted a 4.00 ERA over those 22 starts. If Harvey has an ERA right around 3.50, which is probably a tad high, he will allow approximately eight fewer runs over the course of the season. Last year, Daisuke Matsuzaka and Jenrry Mejia made 16 starts for the Mets, compiling a 4.61 ERA over those appearances. Since those 16 starts would most likely be manned by Montero and Syndergaard, and we conservatively give them an ERA around 4.00, that would be six fewer earned runs allowed. So, just having Harvey return to the rotation and having 16 or so starts from Montero and Syndergaard will create 14 fewer runs allowed already. That would increase the Mets run differential to a positive 25.
The bullpen also can’t be discounted. In 2014, six pitchers made relief appearances for the Mets that will not in 2015: Matsuzaka, Gonzalez Germen, Dana Eveland, Jose Valverde, Kyle Farnsworth and John Lannan. These six pitchers combined to throw 135.2 innings in relief, approximately 28% of the Mets relief innings, and posted a 4.25 ERA. The rest of the bullpen threw 343 innings and had an ERA of 2.70. If you split the difference and say that the 2015 bullpen will have an ERA around 3.50 for those 135.2 innings, that would lead to the team allowing 11 fewer runs in 2015. That pushes the run differential to plus 36.
All of this so far has ignored any offensive statistics, and we’re already close to the magic number of plus 50, at least 88 wins and involvement in the playoff picture. Simplifying statistics again, last season David Wright produced 109 runs (RBI + Runs Scored – Home Runs) over 134 games. Even if Wright reverted back to his other previous season low in 2011, where he produced 107 runs over 102 games and Wright played those same 134 games, his run production would improve to say 130 runs produced, a 21 run increase. That would push the run differential to plus 57 and squarely into the 88 win category.
Last season, Chris Young, Andrew Brown and Josh Satin combined to have 379 plate appearances and produced, using the same down and dirty statistic, 67 runs. In Michael Cuddyer’s absolutely worst professional season, with the Twins in 2008 when he posted a 699 OPS, he produced 63 runs in 279 plate appearances. If he pulled that exact season this year, with the 100 more plate appearances, he would produce 86 runs, 19 more than the Brown, Young and Satin triumvirate. That pushes the Mets run differential to plus 76. Since 2010, seven teams scored anywhere between 66 and 86 more runs than their opponents. Those teams averaged 93 wins.
So, with minimal changes and improvements, this team can potentially score 76 more runs than their opponents and that’s leaving out the 316 plate appearances made by Eric Young Jr., who is no longer on the roster, the impact of Wilmer Flores bat, improvement by Travis d’Arnaud, any improvement by Curtis Granderson and any impact made by John Mayberry Jr. Even if Lucas Duda regresses and Juan Lagares’ BABIP shades back to normal, someone amongst that group should be able to pick up that slack to help even out the differences.
So, here’s the truth of the matter. The Mets will win more than 82 games, all of the numbers lean in that direction without too many ifs, ands or buts thrown into the mix. So instead of 82 and 80, here’s a prediction for you: 88 wins and 74 losses.
I too feel like the BP projection is a little too conservative for this team, but I understand at least the rationale behind it. They come into it expecting a little bit of regression from the bullpen, the back end of the rotation, and probably Lucas Duda at first and Curtis Granderson in the OF. Throw in a mixed bag from Cuddyer with how healthy he can stay, and the black hole that is shortstop, the 82-win number is justifiable. That being said, your 88 win prediction is not outlandish either. I think the real answer lies exactly in the middle, with the Mets being an 85-77 team this year.
To me, the raw W-L prediction is much less interesting than overall rank.
BP is forecasting the Mets to tie with the Cubs for the sixth-best record in the NL. It’ll be interesting to see the development of the Mets and Cubs over the next few years. Joe Maddon, money and hitters versus Terry Collins and pitching.
OMG Brian. Did you just intimate we have to put up with Collins for 3 more years?
Ugh.
Let’s hope Citi Field has plenty of black cats still in attendance.
I love the optimism. I agree that this team could posaibly be that good but two things give me pause. Reliance on a healthy season from an injury prone 36-year-old in the outfield. And the additional runs we’ll give up with an infield comprised of 3 third baseman and a DH.
Agree. The team defense could cost quite a few runs, putting a significant dent in those offensive runs gained.
Good analysis. I am bullish on the Mets this year, especially the bottom of the lineup (6-8 hitters), which should be Granderson, d’Arnoud, & Flores. For the first time in a while, 6-8 won’t be complete black holes in the lineup. Even an average of .250/15/70 from the three of them would be solid, much of which may be dependent on the top of the lineup getting on base at a higher clip.
Let’s Go Mets!!!
Unfortunately, the real record is what the rankings and playoff are judged by, not Pythagorean record and other metrics. Given that no runs have actually scored in 2015 and we have the actual record the Mets in past seasons, are you saying the real record didn’t occur? How does Pythagorean record for seasons past have bearing on this season? The Mets should have had a better record…but didn’t in reality. I get it’s the time in the season where predictions pop up everywhere. It’s fun, but let’s be honest, none of it means anything 3 weeks before ST.
It’s also easy to forget the Mets are, well, the Mets. It’s a team that can’t walk and chew gum at the same time. Strange things can happen. Strange things do happen. I want them to win like every Mets fan, but prepared to accept we might face a real injury like another TJ, or a concussion that ends TdA season. I suspect Grandy won’t produce much more than we’ve seen. I look to where the Mets need to win more…at home and versus the Nats. I struggle to see either improving by vast quantities. We may make up a game or two kore on the Braves, but the Marlins will temper that progress. We already own the Phillies. It’s fine to lump all kinds of numbers together and predict from that, but I want to see just exactly who will lose to make up our extra wins.
For the little value predictions mean at this time of the year, I see the Mets in a 5 game window between 79-84 wins.
If the Mets do go 85-77 does that mean we have to endure another year of Collins? Does 85 wins get you into the playoffs? What a horrible thought! Does Collins get credit for bringing the team closer to the playoffs? Or is it just the law of averages? I’m torn here. I want the Mets to win. But I honestly do not see Collins winning a World Series in Queens. Does the analysis take into consideration the manager? if a good manager is worth 2 to 3 wins that could be the difference between getting in or coming close.
I agree Chris. The Cubs are greatly improved and have a manager now. The Padres are not going to suffer for offense. The Dodgers are a more cohesive team. So where are the Mets going to pick up runs? It’s just an approximation to me. Every year teams alter their 25 man rosters. Some more drastic than others, If the Mets go 82-80 does Collins come back? Geez I hope not!
Great rundown…yes the math is a bit dirty, but the simple fact is that this Mets team, based on everything you said, looks a lot like a 90 win club to me.
Scott your analysis really psyched me up. Wow, 88 wins, a good possibility, it should work.
Then I looked at the article and saw that the Mets keep falling short of the pythagorean record. Twice in a four year span. I would say that the managing isn’t on par with the players abilities. Good managers get more wins out of their players not less.
>> So, here’s the truth of the matter. The Mets will win more than 82 games . . .<<
Scott, you can't logically conclude your analysis by stating "here's the truth," followed by a guess about the future.
Which, you know, hasn't happened yet.
So nobody knows what "the truth" will be, no matter how many numbers they crunch. The only truth is that we don't know.
In regard to win total, the best thing going for the Mets now is that neither the Phillies nor the Braves are trying to win this year: 38 games. Such a weak division.
I know Alderson has been pulling the Pyth out of his well whatever to help ease some fans pain. The truth of the matter is that among the sabr elite Base Runs is considered the Cadillac of the win prediction statistics and better than the Pyth. So why does Alderson rated as the 12th best GM of all time by a couple of his sabr buddies recently not use the preferred method of his friends with pens in high places? Because the Base Run statistic last year actually had us as a 77 win team. One publication recently had us as a 79 win team and now BP with 82, it might be reevaluation time for some among us I have consistently said that if this is all we do this is a 500 team one or two games north or south. I stand by that.