Momentum in Baseball
by mrmetaa on May 21, 2009 2:41 PM EDT
37 comments
Hello,
I am a frequent reader (usually once a week), a sabermetric novice, and a passionate baseball fan.
I recently got into the debate with a very smart baseball friend of mine over the subject of momentum in baseball. We are both stubborn, and we are both persistant in our argument.
My personal opinion is that there is momentum in baseball on the individual and short term level, but over the long run as a team, everything balances and goes back to the point of whether the team is actually talented or not.
My friend goes by the belief that when a player says "the homerun in this inning was a real shot in the arm for the team." He believes momentum is a huge aspect of the game.
I have looked at a few fanposts (most notably one from 2005), online, and through various others sources, but can not find substantial evidence either way.
I was wondering if anyone could provide some insight on the subject. Good articles, arguments, and papers providing either statistical or empirical evidence would be awesome.
Thanks!
0 recs |
37 comments
Email
Print
Comments
Many have found it doesn't matter on an individual basis
For example, The Book says “Knowing a hitter has been in or is in the midst of a hot or cold streak has little predictive value” (pg. 59)
And this article looks at hitting streaks and finds no evidence of the hot hand.
Whether the same holds true on the team level, I don’t know anyone who’s looked at it, but I’d assume that’s case if you hold everything else equal.
The difficulty is that not everything else is held equal. The pitcher may be running out of gas, or may perform worse from the stretch. A hit may cause the manager to bring in a reliever who just doesn’t have it today. Early strikeouts may make hitters more aggressive than they otherwise might be and they might start swinging at pitches further away from their hitting zones.
I’m guessing someone could find a way to attack the problem, but there’s going to be a whole lot of noise involved.
by Dan Turkenkopf on May 21, 2009 3:03 PM EDT reply actions 0 recs
I never beleive in momentum, only "de-momentum"
Here is my threory on it:
1. I think player and teams are generally playing at or 100% of their ability. They are professionals and trying to win.
2. People can’t give 110%, they can only give what they got, not more than that.
3. All a player can do is try less or de-momentum
With this, I think players and whole teams can not be trying that hard and really stay down.
by Jeff Zimmerman (TucsonRoyal) on May 21, 2009 3:05 PM EDT reply actions 0 recs
Basebal is an extremely individual sport.
If players can show momentum, then that’s going to rub off on the team’s overall performance, even if it’s only 1/9 as much.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 21, 2009 3:26 PM EDT reply actions 0 recs
I'm going to have to say yes.
With the 07 Rockies as Ex. A
by bs.uf15bosox9bears23 on May 21, 2009 4:06 PM EDT reply actions 0 recs
Long streaks happen, even for things without momentum.
Someone needs to look and see if there are more long streaks than you’d guess if winning baseball games were probabilitistcally random. And they’re not, considering you can play streaks of bad teams or streaks of home games.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 21, 2009 4:19 PM EDT up reply actions 0 recs
There is some article some where, I think in the "Numbers Game" where every streak in baseball can be explained by random chance except Dimaggio's
I will have to look for it
by Jeff Zimmerman (TucsonRoyal) on May 21, 2009 4:28 PM EDT up reply actions 0 recs
You mean you'd expect a streak of that length to occur given all the years of baseball and random chance?
Cool.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 21, 2009 4:43 PM EDT up reply actions 0 recs
I think it's covered in The Drunkard's Walk by Leonard Mlodinow
I’ll see if I can find the quote later.
by Dan Turkenkopf on May 21, 2009 4:54 PM EDT up reply actions 0 recs
Here it is...
It’s a paper by E.M. Purcell where he quotes (paraphrases?) Stephen Jay Gould, saying except for DiMaggio’s streak, “nothing every happened in baseball above and beyond the frequency predicted by coin-tossing models.”
You can read more about it in Gould’s Streak of Streaks.
by Dan Turkenkopf on May 21, 2009 7:09 PM EDT up reply actions 0 recs
Thanks
I know read it some where. It is a pretty money quote.
by Jeff Zimmerman (TucsonRoyal) on May 22, 2009 12:46 AM EDT up reply actions 0 recs
Poz had a post on that sometime in the last month.
If you were thinking, you wouldn't have thought that.
by Warden11 on May 24, 2009 1:45 AM EDT up reply actions 0 recs
This is from Posnanski:
He even pointed to this very interesting Cornell study which contradicts one of the most commonly believed notions about sports: That Joe DiMaggio’s 56-game hitting streak is a freak occurrence, perhaps the only truly freak occurrence in the history of sports. Stephen Jay Gould, the eminent evolutionary biologist and historian, wrote often about this, saying that while everything else could be explained by the noise of probability, DiMaggio’s streak was so improbable that it simply could not be explained.
But the Cornell folks did a simple simulation of every single baseball season going all the way back to the beginning. They did that simulation 10,000 times. These are people after my own heart.
And here’s what they found: A 56-game hitting streak or longer came up in almost exactly half the simulations. This is a reminder that it all depends how you look at probability. Yes, the probability of Joe DiMaggio getting a hit 56-games in a row is off the charts. But the probability of ANYONE in the long history of baseball getting a hit 56-games in a row — Ty Cobb, Hugh Duffy, Ted Williams, Joe DiMaggio, Duane Kuiper, on and on and on — is not off the charts. It’s closer to 50-50.*
http://joeposnanski.com/JoeBlog/2009/05/08/hotter-than-hot/
If you were thinking, you wouldn't have thought that.
by Warden11 on May 24, 2009 10:28 AM EDT up reply actions 0 recs
For every 07 Rockies..
there are 97 Marlins, 00 Yankees, or 06 Cardinals…all teams that struggled in September but thrived in October.
And then there are 01 Athleticsand 95 Red Sox, teams with higher than .750 winning percentage in september that lost in the first round.
Seems to me that once a team makes the playoffs, there is close to a 50% chance of winning a series regardless of late-season surges.
#269
by mrmetaa on May 21, 2009 4:57 PM EDT up reply actions 0 recs
True.
I wanted to see the response to using the Rockies as proof in this argument.
by bs.uf15bosox9bears23 on May 21, 2009 6:07 PM EDT up reply actions 0 recs
Flip a coin 500 times. About half the times it will land on heads and half the times on tails. But during those 500 flips you’ll get some streaks – 4 or 6 heads in a row, or maybe even something crazy like 11 tails in a row. Is that “proof” that momentum is an important aspect of coin filliping? Of course not. In fact, we know that coin flipping is decided by nothing but chance. It’s just that streaks can happen even for things that are 100% decided by chance. It’s the way probability works.
The mere existence of winning streaks isn’t evidence, let alone proof, of the existence of momentum in baseball (or in any field).
Adoptive parent of Noah Lowry. Because he was awesome once, and, goddammit, he shall be awesome once again!
I hope.
by Cookyman on May 23, 2009 12:29 PM EDT up reply actions 0 recs
The statistics cliche is that...
It’s most unlikely thing is that you don’t observe any unlikely things happening.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 23, 2009 2:22 PM EDT up reply actions 0 recs
but
baseball games aren’t pure chance the way coin flips are. Better teams have a better probability of winning, worse teams have a lesser probability of winning. I don’t really know if that helps or hurts the momentum argument, but it should probably be mentioned.
by cjmulrain on Jun 2, 2009 12:40 PM EDT up reply actions 0 recs
Yes.
If you’re going to test for streaks using a statistical test, you need to account for the fact that you’re using coins with various weightings for each trial. In general, that will make you expect more, longer streaks, a result of teams and players facing easier opponents in a row. There would be more shorter streaks, too, against a series of difficult opponents, but we tend not to notice or care about those.
Good point.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on Jun 2, 2009 1:26 PM EDT up reply actions 0 recs
Let's see...
The Rockies won 22 out of 23 games I believe. Based on their third order record before they started on that streak, I would call them an 86 win team. So what are the chances that an 86 win team wins 22 out of 23. The binomial distributions says that their is less than a 1 in a million chance that will happen. Does that prove that momentum actually had an affect on them winning that many games? No. However, it definitely seems that their was something else in play besides random chance.
St. Louis Cardinals... defying win expectancy since 2008
by vivaelpujols on May 23, 2009 2:37 PM EDT up reply actions 0 recs
Some other possible factors...
- Was their team better in September than it was the rest of the year? (I.e. new players, skipping fifth starters, etc.)
- Did they play more half their games at home?
- Did they player weaker teams or face worse starters and weaker lineups that typical for their opponents?
My guess, although it’s not necessarily true, is that the Rockies would be expected to win more than at an 86-game clip those 23 games.
_ _Next, yes, that streak happening right then was very unlikely. But wouldn’t a 22 out of 23 streak in August been nearly as dramatic and useful for making the post-season? In May?
What if it was 21 out of 23? 20 out of 23? All slightly less dramatic, but we’d still be amazed to a high degree. And 22 out of 23 isn’t exactly 23 out of 23, is it? If we’d seen 23 out of 23, wouldn’t we re-do the math to show just how unlikely that would have been?
One mistake we tend to make with probabilities is drawing the cut-off at exactly what we saw and computing the probability of it happening. That’s a form of selective sampling. The only reason to define that situation is exactly because it happened.
_ _One in a million, huh? That’s based on the 23 game sample picked because of when the streak happened. For one team, there are about 140 days in a season for a 23 game streak to start. For 30 teams, that’s 4200 start games per season. Over 75 years of baseball, that’s 315,000 streak start games. So one might say that seeing a streak like the Rockies had is a one-third chance.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 23, 2009 3:01 PM EDT up reply actions 0 recs
I can't really argue with any of that
About the selective sampling, where should I start it? The whole point of this was whether to see if momentum had an affect. So, I guess a proper way to do it, would be to use a streak that happened in the middle of those 23 games, when they were already momentumized. So let’s say we take the last 17 games of the streak, after they had already ripped off 5 in a row, thus were sufficiently aware enough of the streak so we can asses if it had an impact or not. After that, they won their 16 of their next 17 games.
If you divide the season by 17 games, that gives you 9.5 chances when a streak could start. I choose to separate the steaks as to avoid the chance that they merge, which would obviously make it more statistically likely to win 16 out of 17. So, .095 * 30 * 75 = 213.75. Multiply that by the chance that an 81 win team wins 16 out of 17, and you get just a 2.75% chance of any given team going on that steak. The Rockies were a better than average team, so their chance would be greater, but still pretty unlikely. Again, that doesn’t prove anything, but it makes it likely that their was something else in play.
St. Louis Cardinals... defying win expectancy since 2008
by vivaelpujols on May 24, 2009 2:43 AM EDT up reply actions 0 recs
I think the main point is that you can't really find exact probabilities of streaks, espeically in isolation, because of all the issues I mentioned.
You can’t judge momentum based on one streak. Just like Newton didn’t look at just one object in motion to conclude things about real momentum. You need to look at all streaks and non-streaks in aggregate and see if what we’d call momentum actually predicts the future.
There actually are statistical tests for streakiness. Assume things are random (accounting for stuff like strength of schedule home games, etc.) and figure out how many winning streaks of whatever length you’d expect given all the games that are played. Compare that to what actually happened. Perform math voodoo to quantify any differences.
_ _An example I thought about yesterday was the genetic makeup of any of us. Given all the pairs of people who had to sexually reproduce and all the possible ways their chromosomes could have interacted to produce new DNA codes, the probability of each of us having our exact DNA is infinitessimally small. But given that there are a few billion people alive, we have to have some sort of DNA pattern. Picking the ones that occurred mostly by chance and calling them a probabilistic miracle seems to miss the point, no?
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 24, 2009 9:29 AM EDT up reply actions 0 recs
I'll add that...
I’m not trying to be a downer with any of this commentary (or ever, ideally) and I’m not trying to end any of the discussion. Hopefully any comments I make (and I always hope this about others’ comments) advance the discussion somehow. By “advance”, I don’t just mean go more forward down the same path, but re-direct the path to avoid blockades or take a tangential path that some people can head down that’s also interesting.
For example, there very well could be streakiness in baseball, by either teams or hitters. But any study that tries to detect it needs to be aware of the issues presented. And even if a study can’t detect it, it might still be there.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 24, 2009 1:11 PM EDT up reply actions 0 recs
I wouldn't have the math chops to do a statistical experiment like the one you mentioned
I should agree with you. Normally when I think about baseball, I realize how random it is and how most of the old age “truism” are false. For some reason, I can’t help but think that an extreme streak like this is more than just probabilities.
St. Louis Cardinals... defying win expectancy since 2008
by vivaelpujols on May 24, 2009 2:13 PM EDT up reply actions 0 recs
I took a look at momentum...
…as it relates to how teams play in the postseason. Hint: it’s, uh, not very important.
by cwyers on May 21, 2009 6:48 PM EDT reply actions 0 recs
Thanks Everyone
I just wanted to thank everyone for their responses.
#269
by mrmetaa on May 21, 2009 8:36 PM EDT reply actions 0 recs
How could it not be an important part of baseball.
Look at the Padres who just got their 10 game winning streak snapped. They won all those games on momentum. Same with little league. A team scoring 10 rus in a game by the 5th inning and that is still going stron doesn’t mean they are on a roll, it means they have good momentum going. Some people will disagree, but I say momentum is a big part of baseball.
Rebuild and Restock.
by trademaker on May 27, 2009 8:55 PM EDT reply actions 0 recs
How do you know the Padres won all those games on momentum?
If they’d lost their fifth game of the streak, would that have caused them to play worse in games 6-10?
If a team scores 10 runs by the fifth inning, I’d say the reason they’ll win is because they’ve already scored a lot of runs, not because they’re any more likely to score more runs.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 27, 2009 9:17 PM EDT up reply actions 0 recs
True
You got me there. But you have to agree that when a team is in the pennant race or when they trade for a player on the deadline, is that called “momentum” or simply “just being hot”
Rebuild and Restock.
by trademaker on May 27, 2009 9:19 PM EDT up reply actions 0 recs
I've had the same thoughts...
but I wonder if it doesn’t come from us knowing that momentum does help IN games in little leage/babe ruth/high school/ legion? A lot of us have experienced that feeling from our days and we believe it should work the same in MLB.
If you were thinking, you wouldn't have thought that.
by Warden11 on May 27, 2009 10:13 PM EDT up reply actions 0 recs
The argument against momentum is that players' true talent level do not change based on recent performance
If you think that momentum is like a hot streak, on a team wide level, than you must except that it doesn’t exist as hot streaks have pretty much been proven to have no predictive value.
St. Louis Cardinals... defying win expectancy since 2008
by vivaelpujols on May 27, 2009 10:54 PM EDT up reply actions 0 recs
The argument is with PROFESSIONAL players though,
as a high school coach- I DO believe that 16 and 17 year old kids are influenced by streaks.
If you were thinking, you wouldn't have thought that.
by Warden11 on May 28, 2009 12:02 AM EDT up reply actions 0 recs
Maybe
I play for my high school and I don’t feel that momentum has a big impact on me. When I’m at the plate I just concentrate on try to work the count, and I pretty much shut out what’s happened during the past few games. Baseball isn’t a sport like football, where you can elevate your game with a bunch of adrenaline; baseball takes a calm and repetitive approach to each at bat.
St. Louis Cardinals... defying win expectancy since 2008
by vivaelpujols on May 28, 2009 12:20 AM EDT up reply actions 0 recs
Did this a while back...
… and sent it to my favorite baseball writers at ESPN [Neyer’s rob.neyer@dig.com kept bouncing the emails back :(]
“While watching the Mets game yesterday, I heard [Ron Darling, I think] mention something about expecting a high scoring game after a first inning with six runs scored.
I’m always interested in finding how true ‘conventional wisdom’ type things are, so with 2000-2004 Retrosheet data in hand…
I took the 12000 some-odd games from 2000 to 2004 and grouped them by the number of runs scored in the first inning. I calculated the average runs scored per 8 innings [to adjust for extra inning, weather shortened and games where the home team doesn’t bat in the last inning] for innings 2 onward.
1st RS/9 2nd+
0 8.3 8.3
1 9.4 8.4
2 10.8 8.8
3 11.8 8.8
4 12.8 8.8
5 14.3 9.3
6 15.3 9.3
7 16.4 9.4
8+ 18 8.8
The rate of run scoring after that big first inning doesn’t vary by a whole lot, which would seem to say that a big first inning has little bearing on what will happen in the rest of the game."
by erosen on May 29, 2009 9:11 AM EDT reply actions 0 recs
That's both teams combined, right?
Those plateaus and jumps are interesting.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
by Sky Kalkman on May 29, 2009 9:43 AM EDT up reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
FanShots
Quick hits of video, photos, quotes, chats, links and lists that you find around the web.
Recent FanShots
BtB on Facebook
Certain photos copyright © 2009 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
