with one out and a man on second you can expect to score 0.68 runs in the inning.
with none out and a man on first you can expect to score 0.86 runs in the inning.
In the former situation you ARE more likely to score exactly one run (24% of innings with man on second and one 0ut end with one run scoring, just 18% in the first base, none out scenario). But that’s more than made up for by the chance of scoring bunches of runs.
Scoring 2 runs, man on first none out: 12% of innings; man on second, one out 9%
Scoring 3 runs: 6% vs. 4%
And so on. All in all about 24% of innings end in multiple runs scoring if you have a man on first and none out, and that percentage drops pretty sharply to about 17% if you successfully sacrifice him to second.
So unless you absolutely MUST HAVE one run, or there are other considerations (eg terrible hitter at the plate) you are better off not sacrificing.
Given that the greater run expectancy with 0 outs comes from bunches of runs, I wonder how the data looks between eras where HRs were fewer vs periods like the “Steroid Era”? Also, if HR are down now, vs Stoll’s data set as a whole, does that materially change the run expectancy calculation?
And the Pirates too, they’ve started out strong. I think the Reds are going to sneak out a win at some point to double their win total for the season to date. Bank on it!
See, this is the difference between playing to win a specific game, and playing to score more runs over a full season. The difference is primarily highlighted in “must win” games like playoff series and down the stretch of pennant chases; but really, since every regular season win counts at the same weight as every other in the final tally, sometimes I question why has the latter view has come to dominate.
While you can “expect to score more (total) runs” by not bunting with a man on first and nobody out, as illustrated by the 0.86 versus .68 runs on average over a VERY long period of gathered baseball stats, you’re more likely to score AT LEAST one run by bunting, right? Not just “exactly one run”, but “one or more runs”. There’s nothing in bunting that precludes you from scoring more runs, only that you’ve given up an out that could have been an extra-base hit.
That’s the statistical premise behind the current culture of “don’t-steal, don’t-bunt, advance bases station-to-station and wait for the big swing” baseball. It’s fundamentally driven by the “Pythagorean” model that predicts a team’s win-loss record over a long 162 game season will conform closely to a ratio of total runs scored for the season (squared) versus total runs allowed (squared), so this strategy (a) aims to maximize that numerator, and (b) treats it as an independent value from the denominator.
What I find a problematic with this view is that each win counts the same, whether a 1-0 win or a 10-0 win. Teams that wildly overperform or underperform their “Pythagorean W-L” are often called “lucky” or “unlucky”, because “there’s no such thing as clutch hitting”, but come on. There can certainly be more to it than that.
To use an extreme example, it could be like what happens when I play Scrabble on my iPhone against the AI at “Expert” level, which constantly drops words on me I’d never seen or would not remember, and regularly crushes me by 100 points or so… Yet I have an overall 53% winning percentage, because I place tiles strategically, make an effort to balance my rack for vowels and consonants, and try to save my blank tiles for bingos, instead of always going for the highest scoring word I can play on every turn. My average win margin over the AI is relatively small - under 20 points - but I win more OFTEN than the AI does, even though its victories over me are far more lopsided.
Further, emphasizing the “numerator” of the Pyth W/L relationship does, in fact, affect the denominator, because unlike (say) American football, baseball requires the same players who bat to also take the field. How often do we see GMs pursue a player “with pop” at a position despite below average defense? Maybe their bat does “more than make up” for the defensive shortcomings, but who can really say? Defensive stats are far from clear.
All this is to say - if you trust your pitching and defense to win close games, it seems to me you should be considering the “small ball” approach that maximizes the scoring of AT LEAST ONE RUN once a man is on first base with nobody out, with speed, baserunning guile, and putting the ball in play strategically (swinging the other way, choking up, etc). Because that is equivalent to saying, I expect to hold and win with a one or two run lead.
And that the front office philosophy that says, you can never score enough runs, is what leads to lineups of bat-first, glove-second players who score runs in bunches while also stranding lots of baserunners, allows for massive infield shifts because the Big Lefty Swinger is not going to bunt down the third base line even if the third baseman is ten feet from second base, prompts extremely careful pitching matchups because One Big Swing Could Kill Us, all resulting in 4 hour games.
That’s really what’s slowed the game down, not instant replay or too many intentional walks.
That is kind of the opposite problem; the Pythagorean method doesn’t treat wins and losses at all, but assumes the 10-0 wins translates into two or three wins when figured into the equation.
There is more to it that that, and modern sabermetrics tries to account for that.
First of all, there’s no such thing as a clutch hitting ability. None has been demonstrated, anyway. A player can have amazing or terrible clutch stats in a given period of time; David Ortiz was stupid great in the clutch in 2005, and should really have won the MVP Award. But his clutch performance the rest of his career was pretty much exactly what you’d have expected him to do given how he usually hit.
That said, there is absolutely no doubt whatsoever, none, that if you could space all your runs out evenly, as opposed to scoring them in bunches, you’ll win more games. If you had a choice between having a 50% chance of scoring one run in every inning, or having a 15% chance of scoring four runs in an inning, which would you want? The former will be MUCH better; you’ll win more than half your games. the latter is terrible; you’ll finish last. With the former you’ll almost always score 3-6 runs a game and have a chance to win, but with the latter, which you will score more runs in the season, you’ll win games 8-x and 12-x a lot and get shut out a lot. Your runs shall be wasted.
The obvious problem is sports don’t work that way; you cannot choose to take your grand slam and space the runs out. In most situations, you do not know if a single run is enough to win (in fact, in quite a lot of situations, you should assume it’s not.) There ARE situations where playing for a single run makes sense; to use a super clear example, man on first, no one out, bottom of the ninth in a tie game. Those situations don’t come up much, though, and so the number of times a manager should call for a bunt is really, really small.
This is not to say deviation from Pythagorean record is all luck. Clutch hitting has a lot to do with it - it’s just that clutch hitting is itself mostly luck, not a repeatable skill. There is evidence that the quality of a bullpen contributes to exceeding the Pythagorean projections. What there isn’t evidence of is that small ball helps.
Sure, but you’re no longer talking about Pythagorean records, because, of course, bad fielders affect the denominator. It’s accounted for.
If anything, the value of defense is appreciate more today than it ever has been. The improved, albeit still a work in progress, ability to measure fielding contribution is putting real numbers to what was once subjective. When you see that a really good defensive shortstop or center fielder can be worth the equivalent of 25 homers, that makes you damn interested in fielding.
I don’t know what bat-first, glove-second teams you’re seeing but that philosophy has been dead for five or ten years, anyway. The Houston Astros were the best hitting teams in the AL last year and the Yankees were an excellent defensive team, and the Astros good enough.
.
Gosh, I had to be Mr. Disagreement, but no, it’s not.
Run scoring is not high. It’s not up at all. It’s DOWN from where it was 10, 15, 20 years ago. it has ticked up a little bit since 2014-2015, but is pretty much average by historical standards, sitting at around 4.5 runs/team/game in the AL, 4.2 in the NL (It’s down so far in 2018, but it’s cold, so never mind.) That level is way below the late 90s and 2000s, when run scoring was insane, and in line with the historical average.
What’s up is home runs - but fewer players are reaching base, which of course is the what a sabermetric approach to offense says you shouldn’t do. and which should logically lead to shorter games if we’re just going by the level of offense. However, teams have put a lot of emphasis on better selection and training of pitchers, and are willing to carry glove men again.
So, to answer some of your points, most of which I agree with:
I know that Pyth W/L weights winning 10-0 games more than winning 1-0 games, that’s what I was saying is literally unrealistic, because no, they don’t count for any more wins in the standings. It’s done that way to tell you something about if a team is “powerful” relative to their opponents and thus expected to to better in the long run over a field of similar opponents.
I also agree there’s “no such thing as clutch hitting”, and that “bad fielding affects the denominator” is a mathematical fact, - I was grousing that that SHOULD mean they place more emphasis on defense.
My point is that defense stats are a new art, and that yes, I do see front offices going bat-first, glove -second… So who’s the “They” I’m talking about? At the very least, the team I follow the most closely (the Mets), for who I am a season ticket holder, and to which I attended a Hot Stove Report this past January where I was told, in exactly so many words by Sandy Alderson himself, live and in person, that that was how they had constructed the roster.
That first and foremost, they valued power and OBP (high walks, high strikeouts, but big swings and high slugging percentage) above hitting for average or speed, and that defense was a “nice to have”. (He literally used that exact phrase, in a half-dismissing kind of way.)
It may well be true that this is an approach that MLB teams are trending away from, in general, and that Alderson himself might now prefer a different approach if he had different financial constraints on him (not going to get into that)…
And I know that runs scored has not gone up in baseball overall - but it’s my perception that “trying to score runs in bunches” has gone up over the past 10-15 years, and is not dying down (again, colored by my rage-tinted glasses at my own team’s GM).
You make some interesting points (the great bulk of which I have snipped away, sorry :)), but the thing is–the bolded part of your post is wrong. (Boy, am I ever feeling like snopes in this thread…)
You are in fact slightly MORE likely to score (at least one run) if you DON’T bunt.
Yeah, I was a little surprised too, and I know it seems kind of weird, but them’s the stats.
According to my link, with a man on first and no outs, teams fail to score in about 57.65% of innings. Thus, they score at least once in about 42.35% of innings.
Whereas with a man on second and one out, teams fail to score in about 59.45% of innings. So, in this situation teams score at least once in about 40.55% of innings.
That’s over hundreds of thousands of innings. There’s no cherry picking of the data going on.
So: If you don’t bunt, you come out ahead in two ways. First, you are simply more likely to score, period. It’s not a big difference–two out of every hundred innings–but it’s there. Second, a chunk of one-run innings become two-or-more-run innings.
According to this information, you can argue that even in the most obvious bunting situation (mentioned by RickJay, tie game, man on first, none out, last of the ninth, a scenario in which one run wins the game and no other runs are necessary) you still shouldn’t bunt, because you have a 40.55% chance of winning in that inning if you sacrifice successfully but a 42.35% chance of winning in that inning if you try something else. Might as well play for a bigger inning even if you don’t need it, because that strategy gives you a better chance of scoring the winning run.
Obviously, the specific odds are going to shift somewhat based on how good a bunter the guy at the plate is, how good a hitter he is, how strong the defense is against the bunt, the element of surprise, the value of showing bunt to draw the infielders in, etc, etc, etc. But* in genera*l, the evidence is clear: a deliberate sac bunt is going to make you less likely to score in that inning, and isn’t therefore a winning strategy.
Yanks have a crazy one going against Baltimore. 13th inning so far. Job saved the gave stealing a homer in the 13th. CC had to leave early with a hip issue, very scary. DD hit a big game tying homer in the 8th. There was a strange double play that didn’t happen. Bullpens are going strong.
I’m happy that the USA women won the Olympic gold medal in ice hockey. However, the Chicago White Sox don’t need to spend an entire inning interviewing one of the players while ignoring the game on the field, the jerks aren’t even showing the game most of the time, just focusing on the announcers!
Isn’t Jason Heyward in the lineup because of his glove? That was…not good. A badly misplayed ball off the wall and now the Brewers are threatening. The Cubs really need to win this series and get back on track.
If MLB is going to make rule changes, how about a rule that declares any game a tie that goes 14 innings? It’s ridiculous to play the equivalent of 2 complete games (if it goes longer than that) just to avoid a tie score. Personally I’d make it 12 innings but that would be an even harder sell.
Astros-Padres game ends with a bizarre play; 0-0, bottom of 10th inning, runner on second, two out, the (Astros) batter hits a high shallow pop-up, the first baseman comes charging in to catch it, overruns, the ball drops behind him, and the runner scores. Walk-off 20-foot single.
Nah, fuck that. You want a man on base, get him there yourself. I just see nothing wrong with a game ending in a tie. And I freaking HATE extra inning games!
If you have to break a tie, make the managers run the bases for time. Fastest manager wins the game. They’re the only coaches in uniform on the filed of play - make them a definitive factor in event of a tie.
As odd as it seems, I’d be fine with a tie rule after a certain number of innings with the caveat that the game has to be made up if it will end up mattering in the standings at the end of the season. No team wants to blow through their entire bullpen and it always seems like the long extra inning games occur on a night game before a day game the next day with no time to get a minor league call up.
Most fans love extra inning games and, let’s be honest, overly long games are rare occurrences. A team can go a whole season without playing a game longer than 12 or 13 innings.
