I was struck yesterday by the fact that both the Cavs and Rockets were up at the half but lost the game. I wasn't watching the Cavs game - but the Rockets (specifically Tmac) looked dominant over the Mavs in the first half and then just petered out. I wondered how good a predictor a half time lead was for eventual victory.
I have only looked at the games that have been played in the last 7 days - 51 in all.
(1) Of those one team was leading at the half in 49 of them.
(2) That team won in 32/49 of the games (65%) of the games. This number is not significantly above chance - but if I looked at the whole season I bet that it would be.
(3) Of course, some proportion of those games (Cavs v. Suns for example) were blow-outs from the very beginning, so I removed those games where a team was up by at least 15 points at the half (which happened 6 times and in all 6 cases the team with the lead went on to win). Now 26/43 of the times (60%) when the team was leading at half they won. This is not significant, and even if it was it means 40% of the time when a team was losing by 1-14 points at the half, the team came back and won.
Maybe all of those come from behind victories were when the other team lead by only one or two points. Not so!
When the team was leading by 1-4 points: They won 11/18 times
When the team was leading by 5-9 points: They won 11/18 times
When the team was leading by 10-14 points: They won 4/7 times.
Perhaps this week is an anomaly - if I look back over the course of the season maybe a half time lead is a really good predictor - but right now it looks like going into the half with even a healthy lead barely gives you an edge in the final result. (in contrast - for the first three weeks of the football season this year the team with the lead at the half won 80% of the time)
In other news: here is a web site full of rainbows puking.
http://www.rainbowpuke.com/
Subscribe to:
Post Comments (Atom)
2 comments:
I read this post while I was in "Seminar in Electronic Music Performance" this morning. I am seriously into this blog!! Good statistical work -- it would be interesting to see, team by team, which teams are more likely to hold a halftime lead. Do the spurs win 70% of the games in which they lead at halftime? Do the wizards win 30%? Those teams are what I would call a "good" defensive team and a "ridiculously incapable" defensive team. The pistons, even in their current state, I still consider to be a "good" defensive team, however, they also fall into the category "not giving a fuck about 1/2 the teams in the league" and therefore i would not expect their halftime lead retention to be especially impressive. Ch-Cheers.
I agree with Jake's complication. For instance, I have an anecdotal sense that the Cavs do this kind of thing pretty often - lose games that seemed to be in hand, even if the lead wasn't huge, at halftime.
In the Cavs case, commentators tend to chalk this up to "an inability to put teams away." There may be just such a psychological force at work, though it seems to me that depth is the more vivid matrix. The Cavs get a so-so number of bench points, and what they do get comes quite often from 3-point shooters. Late in the game, when starters may be tired or in foul trouble, outside specialists are a low-percentage bet for putting a game away. They may be more effective at piling up quick points than big men when they're on, but when they're not even a weak team like Seattle can get right back in the game, as yesterday.
This is precisely why I am envious of the Pistons' depth up front. It gives them great late-game options for predictable, high-percentage points inside, at moments when the thing needed is DAMAGE CONTROL, the preservation of a medium-sized lead, rather than the spark for a comeback. Not to mention the generally more intimidating defensive presence of a 6-10 forward as compared to a twiggy guard.
I'd like to see some hard numbers on not-giving-a-fuckness at some point as well.
Post a Comment