PDA

View Full Version : Bill James on "When is a game out of reach?"



Lauderdevil
03-17-2008, 09:58 AM
Bill James has an interesting article up on Slate with a mathematical formula re when you can calculate with certainty that a lead in a college basketball game is "safe" -- i.e., there's mathematical certainty that the trailing team cannot win.

He tracks some of the great comebacks in college basketball history (including, e.g., Villanova's win over LSU this year after trailing by 15 with 2:59 to go) and shows that LSU's lead was not yet "safe."

In fact, the single, solitary example he's able to find of a game in which a "safe" lead (per his formula) was blown was -- ugh -- the 1974 game in which Duke led Carolina 86-78 with 17 seconds to go, and UNC came back to tie and win in overtime.

Good reading.

The Lead Is Safe
How to tell when a college basketball game is out of reach.
http://www.slate.com/id/2185975/?from=rss

billybreen
03-17-2008, 10:03 AM
Bill James has an interesting article up on Slate with a mathematical formula re when you can calculate with certainty that a lead in a college basketball game is "safe" -- i.e., there's mathematical certainty that the trailing team cannot win.

He tracks some of the great comebacks in college basketball history (including, e.g., Villanova's win over LSU this year after trailing by 15 with 2:59 to go) and shows that LSU's lead was not yet "safe."

In fact, the single, solitary example he's able to find of a game in which a "safe" lead (per his formula) was blown was -- ugh -- the 1974 game in which Duke led Carolina 86-78 with 17 seconds to go, and UNC came back to tie and win in overtime.

Good reading.

The Lead Is Safe
How to tell when a college basketball game is out of reach.
http://www.slate.com/id/2185975/?from=rss

By my calculations, Maryland's lead in the Gone in 54 Seconds game was 78% safe.

dukelifer
03-17-2008, 10:29 AM
need to look at it more closely- but I ran the case of a 9 point lead with 40 seconds with the ball. That is apparently 100% safe- but a 7 point lead with 40 seconds with the ball is only 51% safe. It is hard to believe that teams have not come back from a 9 point deficit with 40 seconds to go. You foul, put the team on the line- lets say for a 1 and 1 then they miss you get a fast three off in 8 seconds and get to 6 with 32 seconds left- at that point the team with the lead and the ball only has a 38% safe lead. So something does not compute here. In 8 seconds the game went from 100% safe to 38% safe. Now you foul again an the team makes 1 out of 2. So the trailing team is down 7 - again hits a three is down 4 with 20 seconds to go. The lead is now only 11% safe. So, in the end it looks like no lead is really safe unless there is huge margin for safety. While 9 may be technically safe- it is not by a large enough margin. I would expect that 29 or maybe 19 with 40 seconds is safe but 9 does not seem safe although his formula deems it so. So he needs to bring in something else here.

Lauderdevil
03-17-2008, 10:43 AM
need to look at it more closely- but I ran the case of a 9 point lead with 40 seconds with the ball. That is apparently 100% safe- but a 7 point lead with 40 seconds with the ball is only 51% safe. It is hard to believe that teams have not come back from a 9 point deficit with 40 seconds to go. You foul, put the team on the line- lets say for a 1 and 1 then they miss you get a fast three off in 8 seconds and get to 6 with 32 seconds left- at that point the team with the lead and the ball only has a 38% safe lead. So something does not compute here. In 8 seconds the game went from 100% safe to 38% safe. Now you foul again an the team makes 1 out of 2. So the trailing team is down 7 - again hits a three is down 4 with 20 seconds to go. The lead is now only 11% safe. So, in the end it looks like no lead is really safe unless there is huge margin for safety. While 9 may be technically safe- it is not by a large enough margin. I would expect that 29 or maybe 19 with 40 seconds is safe but 9 does not seem safe although his formula deems it so. So he needs to bring in something else here.

James -- who's probably the most influential statistician ever in sports -- addresses this in his article as follows:


Once a lead is safe, it's permanently safe, even if the score tightens up. You're down 17 with three to play; you can make a little run, maybe cut it to 8 with 1:41 to play. The lead, if it was once safe, remains safe. The theory of a safe lead is that to overcome it requires a series of events so improbable as to be essentially impossible. If the "dead" team pulls back over the safety line, that just means that they got some part of the impossible sequence—not that they have a meaningful chance to run the whole thing.

DukeDevil
03-17-2008, 10:45 AM
what about being down 10 with 54 seconds left?

Bluedog
03-17-2008, 10:49 AM
what about being down 10 with 54 seconds left?

See the second post in this thread.

hurleyfor3
03-17-2008, 10:56 AM
In the 1995 unc game in Cameron, we were down eight with :17 to go in regulation as well. Like us, unc in '74 only forced overtime. Of course, they didn't have three-point field goals in '74.

hurleyfor3
03-17-2008, 11:02 AM
By my calculations, Maryland's lead in the Gone in 54 Seconds game was 78% safe.

More explicitly, we had the ball with :54 remaining.

In that game, IIRC Maryland was up 12 with something like 1:10 AND the ball. That lead would be 100% safe, even without the ball.

DukeDevil
03-17-2008, 11:02 AM
oops, don't know how I missed that.

I think they were technically up 10 with 60 seconds left on their last shot if you want to get picky, so that drops the percentage to 70%. Though with each second ticking off the clock till 54 seconds when we scored, the percentage slowly increased to 78%. As soon as J-Will hit that drive it dropped to 38%.

dukelifer
03-17-2008, 12:12 PM
James -- who's probably the most influential statistician ever in sports -- addresses this in his article as follows:

The example that I gave did not seem that improbable, particularly against a poor free throw shooting team - like a Memphis or Clemson (when they don;t play Duke) to get a 9 point deficit with 40 sec to 4 point deficit with 20 seconds left - but I agree that certain things have to happen. But I believe he has looked at this more that I have (about five minutes)- but I find it a bit hard to believe that once you cross the safe line- all it takes is a series of improbable events to lead to victory- but perhaps it is true. Intriguing nonetheless.

billybreen
03-17-2008, 12:23 PM
More explicitly, we had the ball with :54 remaining.

In that game, IIRC Maryland was up 12 with something like 1:10 AND the ball. That lead would be 100% safe, even without the ball.

Interesting. I didn't think to walk back to see if the lead was safer earlier.

EarlJam
03-17-2008, 12:33 PM
I found another example of when a "safe lead" was blown. Well, almost. It should have been.

I typed in a four point lead with two seconds remaining (sound familiar to anyone?)

It said the lead is 100 percent safe.

J-Will for the three! Bottoms! And he's fouled on the play! (Indiana, 2002).

Unfortunately, J-Will missed the free throw and Boozer got hacked and missed the follow up. Former Atlanta Brave Bruce Benedict didn't blow the whistle and he is a stinky face to this day for not doing so. The Monster had to be withheld from his cage. Chaos ensued.

-EarlJam

rsvman
03-17-2008, 03:09 PM
In the bad old days (before the 3-point shot), people used to talk about something called a "differential" and the predictability of a win using it. The differential was the lead minus the time left in the game. If less than 5 minutes remained, and the differential was 5 or greater, "they" said the leading team had a 75-80% chance of winning the game. Obviously, not "safe," but a heck of a lot easier to calculate.

I have always wanted to test the hypothesis using about, oh, 500 games or so and see if there is any truth to it, or how much the 3-point shot has changed things.

IStillHateJimBain
03-17-2008, 03:30 PM
I know I read about this somewhere but I can't find any record of it now:

There was a junior college game in Idaho or Montana in the 1980s where a team lost after leading by 7 points with four seconds to play. During a deadball situation, the celebrating players dumped water over the coach's head. It got all over the floor -- the water, that is -- delaying the game and the team was assessed two technical fouls. The other team hit all four free throws, inbounded the ball and hit a 3-pointer to send the game to overtime, where they won.
If someone can find the record of this, I'd like to know what the schools were.

Bostondevil
04-11-2008, 02:01 PM
I enjoyed this article despite the one exception to the rule.
http://www.slate.com/id/2185975/

Is there a tape of that game?

(I don't want to see it, mind, but I don't really remember if the dribble off the foot or the pass off the rim happened first. I was 9.)

Bostondevil
04-11-2008, 02:43 PM
Oops, sorry, I duplicated a thread, don't know how I missed it the first time around. Still, does anybody remember which came first? Or have we all tried to forget?

Bob Ryan wrote a column last week comparing Hansblahblah's game to Kupchak's. Fair assessment I suppose, but I still hate Kupchak much more than I'll ever be able to hate Hansbrough.