A .333 Avg is getting 13.5 hits a game. If you want to factor in walks, teams averaged 3.28 walks/game the last 5 years. Factoring that in, and Team 333 has a 0.383 OBP. This is going to crush the opposition. Opposing pitchers are going to have a 1.86 WHIP. The RC formula spits out 838 runs on the year. They’re going to the playoffs if they have even average pitching.
To me, this seems a problem best solved with some type of Monte Carlo simulation, but you’d also need a stack of other statistics to go with it (like chances of a double play, how often does someone score from second, how often does a runner advance to third from first on a hit, etc.) You can get quite granular with the level of statistical detail you put into it, though at some point, there’d be diminishing returns. At the first order, you could just have a simple rule like “all runners from second or third score on a hit; all other runners move up one base” and then add complexity as needed. (Though in that scenario, you would never have anyone on third, unless we include walks or the statistics for how often a runner on second only makes it to third on a hit.)
To that point, the leaguewide OBP last season was .317; it’s consistently been around .320 for the past decade. The team with the best OBP last year was the Astros, at .339; they also had the best batting average in MLB, at .267.
Sources:
https://www.baseball-reference.com/leagues/majors/2021-standard-batting.shtml
In modern offensive terms. 838 runs is a lot of runs. Granted, you could be unlucky like the Blue Jays last year and be pretty good and still miss the playoffs. A bit harder with one more team in to miss; given last year’s AL average of 745 runs scored, your team should go about 90-72.
Having said that, I think such a team might actually do better than 90-72. We are assuming here that the arrangement of runs scored would be relatively normal, but that might not be true. Team .333 would, I suspect, be much more consistent in how many runs it scored per game; a steady stream of hits and walks might mean fewer huge blowout wins but also fewer games where they score 0 or 1 runs. Consistently scoring 5-6 runs a game as opposed to significant swings will actually win slightly MORE games that the Pythagorean expectation accounts for. (To see what I mean, imagine a team that averages 10 runs a game, but always scored either 20 runs or no runs; they cannot do better than .500 and might even lose a game 22-20.)
Even with all that taken into account, wouldn’t it depend on who the batter is? Sure, Barry Bonds, Mark McGwire, or Sammy Sosa might be justified in swinging for the fences every time up. Even when they come up short, they could at least be hitting a double. But what about those contact hitters that you mentioned? Would the team as a whole be better off if Ichiro was swinging for the fences very time he comes up?
Every batter is different and has different abilities, so it depends how they’re trained, how strong they are, stuff like that. Obviously, Ichiro was an outstanding hitter despite not hitting many homers or drawing many walks, but Ichiro was an outlier.
That said, no one has a problem with you hitting for average; if you hit .372 like Ichiro did one year, no one will care if you hit homers. “Small ball” usually means things like bunting and stolen bases, and the evidence is very, very clear that it doesn’t work - it NEVER worked, really. Except in very rare, specific circumstances, sacrifice bunts are insanely stupid, and stolen bases only matter if your success rate is really high.
These stats should be readily available, right? And I think the three stats you call out are all that we really need for a MC simulation (assuming we are ruling out stolen bases, balks, HBP, etc.).
Agreed. Even in the current MLB environment, a really good contact hitter will still very likely be a really good contact hitter, just as the guys who are the best power hitters will still be among the best at that.
What I suspect is happening is that many of the guys who aren’t in either of those camps, have, over the past few years, changed their swings to develop a higher “launch angle,” and hit the ball in the air more. The stats have shown that ground balls are increasingly likely to wind up being outs, and so, “everyday” players are likely adapting their hitting style accordingly.
A couple of cites:
As a fan of the Royals, the 2014 and 2015 teams that went to the World Series were a variation of the small ball approach. The team liked to refer to it as “keep the line moving”. They had very smart, effective base stealers - few players were given the green light who didn’t have a high success rate. A premium was placed on smart base running when the ball was hit (Lorenzo Cain scoring from first in the 2015 ALCS Game 6 being the culmination of that philosophy). There was far too much bunting and sacrificing though. Giving up outs is so dumb.
I’m also a Royals fan, and I agree with everything you say. In 2015, KC was second in the league in batting average and total hits, but just sixth in runs scored, and second from the bottom in home runs. Royals offense was above average, but they don’t win the WS without their killer bullpen.
Also, it was really hard for the models to figure out what to do with those teams. Everyone vastly understand-rates their win projections. Defense, smart baserunning and relief pitchers are really hard to quantify.
And even then it was only a lot of bunts by modern standards.
How well do the runs scored formulas hold up at the tail ends of the distributions?
I am not sure overall and I’m not even smart enough to be 100% sure what your question is asking, but I think you mean extreme cases, e.g. the best and worst teams. It holds up very well at all levels for team performance, however, simply because MLB teams don’t have THAT great a range of scoring performance. It never happens that you have one team that bats .170 with 50 homers a year and another that bats .350 with 400 homers; if the BEST offensive team scores 20 percent more than the average team, that is a hell of a year.
Just for fun - this proves nothing - I applied the basic formula to the 1997 Blue Jays, who were one of the worst hitting teams of all time, and the 2015 Blue Jays, one of the best hitting teams of all time. It was within one run of their 1997 total - 655 versus 654 - and in 2015 it was low by 41, a slightly larger than normal miss but not terrible.
I ran a bunch of other teams too and it was always pretty accurate - except in the case of the 2021 Rays. RC believes the Rays scored a HUNDRED more runs than would be expected by how they actually hit. I admit I can believe this; I still don’t understand how the 2021 Rays scored so many runs.
So, for comparison, what if we had a batter with loads of power, but poor control? Whenever he hits at all, it’s ALWAYS a homer, but he has a low batting average. How low would his average have to be for him to be equally valuable as the 333-all-singles batter?
He’d have to hit .167 to equal the RC of the .333 singles hitter.
Not necessarily best and worst teams although those fall under what I was asking. But like, does it hold up for teams that get on base but don’t hit for power? Those that hit for power but strike out a lot. It’s possible the formulas is accurate for an average team but falls apart for edge cases like this thread is asking.
From your post and post #7 it sounds like the formula holds up fairly well although only hitting singles and never walking is surely weirder than any season a team has actually played.
Pretty much. It works for teams in the dead ball era, with small adjustments as I noted upthread. Back then there were very, very few home runs.
I just want to thank you all for the discussion. This has been most interesting.
What about a team that only gets walks?
(I’m kidding!)