tag:blogger.com,1999:blog-2893512317902577458.post1139619869364226911..comments2023-10-16T01:03:51.515-04:00Comments on Camden Depot: O's Need More From Struggling Offense In Remaining 12 GamesJon Shepherdhttp://www.blogger.com/profile/03521809778977098687noreply@blogger.comBlogger10125tag:blogger.com,1999:blog-2893512317902577458.post-18625954467967469342016-09-22T16:53:38.477-04:002016-09-22T16:53:38.477-04:00Re: Matt P
You might be right that the high-varia...Re: Matt P<br /><br />You might be right that the high-variance Team A could win more games than Team B given those parameters... but the two situations are actually different:<br /><br />* In my example, both teams score more runs than the average team (~4.5 runs/game = ~729 runs/season). Under this scenario, you would probably want lower variance to reduce the chance that the "0 runs" scenario would come up, as the extra runs in the games that you're already winning are essentially wasted.<br /><br />* In your example, both teams are below average in terms of offense (their expected win-loss percentage is ~32.2%). In that case, Team B will do worse because it's consistently bad (i.e. 3 runs will lose more than 50% of the games), while the high variance of Team A is basically akin to gerrymandering (giving up half of the games to maximize the probability of winning the other half)... but their win probability will still be lower than 50%.<br /><br /><br />In other words: <b>High variance has the effect of regressing a team's win probability toward 0.500</b>. That's good if you're below average in terms of offensive output, but bad if you're above average. So, let's look at the Orioles: They're currently scoring 4.64 runs per game, versus the MLB average of 4.48. That makes them slightly above average, which means that high variance will probably hurt them slightly...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-57355951511476550442016-09-22T11:27:27.748-04:002016-09-22T11:27:27.748-04:00In general, power-hitting offenses are a better be...In general, power-hitting offenses are a better bet for postseason success than non-power-offenses. In the postseason, generally the opponents have better-than-average pitching and defense. All types of offense are reduced, but reducing the number of home runs hit has a smaller effect on runs scored than reducing the number of other hits. So, the Orioles offense is about as well-designed for postseason success as an offense OF THAT QUALITY can be. Anonymoushttps://www.blogger.com/profile/18347209822215185644noreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-34526196746685496112016-09-22T10:53:49.671-04:002016-09-22T10:53:49.671-04:00THIS team is significantly outperforming their Pyt...THIS team is significantly outperforming their Pythagorean expectations.Rogernoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-71208807413126956912016-09-22T08:43:14.422-04:002016-09-22T08:43:14.422-04:00Here, suppose we changed your example, just to sho...Here, suppose we changed your example, just to show why such an approach won't work.<br /><br />* Team A scores 6 runs 50% of the time and 0 runs 50% of the time. This team scores 486 runs.<br /><br />* Team B scores 3 runs 100% of the time. This team also scores 486 runs.<br /><br />Presuming average pitching, Team A will probably win more games than Team B. In the worst case scenario, Team A will certainly do better in comparison to Team B in my case than in your case. <br /><br />This is because after a certain point, there's little benefit to scoring more runs. Whether you score 10 or 20 doesn't matter because you'll probably win either way. Therefore, a team might pay $10M for runs 1-10 in a must-win game, but might only $15K for runs 11-20 (presuming they were the only team that could buy runs). Each run a team scores has lower expected worth than the previous one. That's why you shouldn't be using Pythag in this case.<br /><br />Now, if teams did act like they do in your scenario, then this might be ok. But they don't. So, in order to accurately measure performance, you'd need to do a more sophisticated analysis. <br />Matt Pnoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-31252884309313200542016-09-22T00:00:00.797-04:002016-09-22T00:00:00.797-04:00Re: Dustin
I'd argue that it's actually t...Re: Dustin<br /><br />I'd argue that it's actually the opposite-- a low-variance team might actually be better for the postseason. Consider the following two extremes:<br /><br />* Team A: Scores 20 runs 50% of the time and 0 runs 50% of the time<br />* Team B: Scores 5 runs 100% of the time<br /><br />Here, we can expect Team B to have a better win-loss record than Team A despite having only half the offensive output on the aggregate. After all, 5 runs will allow Team B to win more than half of their games (the league average is about 4.5; applying the Pythagorean expectation formula for the entire regular season would yield ~54.8%). In contrast, Team A is likely to end up with a 0.500 season (since scoring 20 runs will almost always win, while scoring zero will always lose).<br /><br /><br />Now, you could legitimately argue that this is a long-term expectation and that anything can happen in the short term. But the postseason still requires the coin flip to land on the correct side 3+ out of 5 times or 4+ out of 7 times; Team A would have a ~50% chance of winning each series, which yields a <b>~12.5%</b> chance of winning the World Series. In contrast, Team B would have a ~58.9% probability for winning a 5-game series and ~60.4% for winning a 7-game series. This yields a <b>~21.5%</b> chance of winning the World Series.<br /><br />If I have to choose, I think I'd rather be Team B...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-3808704042589314402016-09-21T17:03:24.010-04:002016-09-21T17:03:24.010-04:00I vehemently disagree with the notion that a team ...I vehemently disagree with the notion that a team built like Baltimore is ill-equipped for the postseason. Being built on free swingers who strike out a lot but also hit a whole lot of home runs is probably the ideal way of building a postseason team. The modern adage is the team that wins the World Series is the team that 'gets hot at the right time'. Well, wouldn't the most high variance teams be most capable of putting together a string of games that would be most likely to win these games?<br /><br />What I mean by this is... The Orioles strike out and hit home runs more than any other team. What that leads to is a lot of innings where the Orioles go down 1-2-3 with 2 strikeouts. But it also means that the Orioles are more capable of putting a run up on the board with no runners on base and 2 outs. In a one game sample, the Orioles seem to be the most high variance team. They're most likely out of all 30 teams to score 0 or 1 runs. They're the most likely tout of all 30 teams o score 7 or 8 runs.<br /><br />So, when you're in the playoffs, and trying to win 11 out of 19 games, would you want your team to be MORE high variance than not? It's a series of 19 1-game samples. Losing 7-0 counts as much as losing 1-0. Winning 1-0 counts as much as winning 9-1. The higher variance your team is, the more likely, it seems, that you can piece together a 'hot streak' and win the whole thing.<br /><br />I may be crazy, here, but every team needs to 'get hot' to win a title, and if the Orioles 'hot' and 'cold' are more extreme than other teams, then wouldn't that be a perk and not a bug?Dustinnoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-81954638283640027122016-09-21T15:31:36.901-04:002016-09-21T15:31:36.901-04:00As myself and others have said since the beginning...As myself and others have said since the beginning of the year, this offense isn't built for postseason play. And the back-to-back Trump/Davis combination in the middle of the lineup is an eyesore. Two players who strikeout a lot and only spray homeruns here and there is not a good recipe for success. The Davis signing is looking worse and worse as the year goes on.<br /><br />Hopefully, the O's can look to signing players this offseason who offer more than a HR for about a quarter of the games. Guys like Bourne rather than Trumbo who can play sharper defense, hustle around bases, move base runners over....all for a much cheaper price. They are already stuck with Davis. Quit adding more guys like him. AKSnoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-16819615588986475042016-09-21T12:00:11.152-04:002016-09-21T12:00:11.152-04:00even if this team lineup stays mostly intact next ...even if this team lineup stays mostly intact next year we are in for the same old same old. until they get some quality starting pitching to help keep the arms in the bullpen fresher you will continue to see a fade at the end when teams are able to add quality parts to their rosters at the trade deadline. with a completely bare farm system for the most part trades for quality are a moot point and no quality free agents seem to want to come to Baltimore. this team unfortunately is going to have to be torn down and rebuilt keeping a few key parts and trading for draft picks AND get someone who can actually evaluate talent to run the farm system cause they sure can't as is. jmo.tony2302https://www.blogger.com/profile/14464582714647079915noreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-64458806123024493132016-09-20T20:37:13.142-04:002016-09-20T20:37:13.142-04:00Good thing Buck listened to me and put Mancini in....Good thing Buck listened to me and put Mancini in.... :-)Rogernoreply@blogger.comtag:blogger.com,1999:blog-2893512317902577458.post-57417497716278095262016-09-20T13:04:39.456-04:002016-09-20T13:04:39.456-04:00Sure would like to see them give Mancini a chance ...Sure would like to see them give Mancini a chance to DH or even play 1st for Davis against a lefty.Rogernoreply@blogger.com