APBRmetrics

I believe a simple and sound idea would deliver each of the following to EVERY college basketball game (variations would work at other levels of play, too):

-ELIMINATION OF REPEATED, DELIBERATE, LATE-GAME FOULING BY TRAILING TEAMS (In my view, this is the holy grail of improvements for the great game of basketball)
-Elimination of late-game deliberate slow play by leading teams
-Guarantee of a walk-off made basket
-Elimination of late-game clock malfunctions, errors, and reviews
-A fair, but not fluky, chance for trailing teams to mount a late comeback
-Elimination of overtime (and the added actual time, increased foulouts, and anticlimactic endings that come with it)
-Reduction in variability of length of games (in actual time)

Without a doubt, the presence of the game clock directly leads to an unpalatable style of play (fouling, four corners, desperate heaves at the basket, etc.) during the final few minutes of nearly any game that’s remotely competitive. This will always be the case (no matter what bandaid rule changes are implemented) as long as basketball determines a winner on a which-team-is-leading-when-time-expires? basis. In order for basketball’s late-game style of play to resemble, well, basketball, the sport must determine a winner on a which-team-is-first-to-reach-a-certain-point-total? basis. With that in mind, I believe the game clock should be abandoned for those last few “minutes” of each game.

This particular format would replace the last four minutes of game-clock-focused play with a comparable amount of game-clock-free (and basketball-focused!) play:
1. Teams play a timed 20:00 first period
2. Teams play a timed 16:00 second period (I would prefer for this period to end naturally - instead of sounding a buzzer and stopping play as soon as the clock reaches zero, allow play to continue until the next whistle)
3. Upon returning from that timeout, the game clock is abandoned for good and the target score is established for that game (Target Score = Leading Team's Score + 7). For example, if the score is 70-62 after 36 minutes of play, the target score would be 77; if the score is 65-64 after 36 minutes of play, the target score would be 72.
4. Teams play the final untimed portion of the game until one team’s score matches or exceeds the target score

Certain adjustments could be made (the juncture at which to abandon the game clock, establishing the target score by adding a flat number other than seven, establishing the target score by adding a percentage of scoring from that game, whether to consider time elapsed in excess of 16:00 during second period when establishing target score, using a natural or buzzer-prompted transition from timed portion to untimed portion, etc.), but the idea is to abandon the game clock just before it would compel a team to deviate from the basic objectives of basketball.

I began exploring the necessity, soundness, feasibility, and superiority of this format (and other variations) in 2007, and I believe in the idea now more than ever. I would appreciate any thoughts (including negative ones) you might have on its viability.

I pretty much like the idea. You don't address the shot clock: do you keep it operating as usual? If not, then the assertion that we'd have Reduction in variability of length of games (in actual time) may be false.
Teams could revert to 1952 and just play with the ball indefinitely; though I don't know why they would, and maybe it would never be a viable strategy.

Great question. The shot clock would remain in play throughout the entire game (and for the reasons you mention)

I like the idea, for the sake of a better game, however I don't like it from the point of view of statistical analysis. The game would cease to be (near) stochastic and all sorts of complexity in tactics would arise.

Well, the bugaboo in statistics is largely about 'garbage time' at the end of blowout games, and this proposal might actually alleviate some of that.

If the final 4 or 5 minutes is replaced by a final 7 to 13 points, then it's more like the 9th inning of a baseball game. Your team might be down 6-0, or 90-75, but you have some chance to win, because time is not also your opponent.

In other words, your chances of going on a 22-6 run are better if you don't have to do it in 5 minutes or less.

Can anybody point me towards any work done on analysing the features/existence of 'garbage time'? It's not that I don't believe in it, more that I think it's probably only a very small blot on an otherwise homogeneous landscape.

Garbage time is typically based on when the game has been decided (win probability > X%).

An article by Bill James on the question: http://www.slate.com/articles/sports/sp ... ingle.html

Statsheet tracks "lead safeness" for NCAA games.

Here is an old APBR discussion: http://godismyjudgeok.com/DStats/APBRme ... =1462.html which references this: http://godismyjudgeok.com/DStats/APBRme ... t=586.html , http://godismyjudgeok.com/DStats/APBRme ... rt=15.html , and http://godismyjudgeok.com/DStats/APBRme ... rt=30.html

Suppose a game is 90-75, with 4 (or 5) minutes left.
Under the current format, the game is well in hand, and the benches will be emptied. The game will be finished by players who seldom play when the outcome is in doubt.
Under the new rules, the trailing team will keep their best players on the floor, and the leading team may follow suit. They want to finish the game and go home.

Under current rules, with the clock running out on the losing team, the game that was "over" much earlier may end as a 5 point margin or 25 points. It's the same game, essentially, but the "garbage time" play of scrubs made it seem much better or worse than it "really" was.

When the end of the game is more like a horse race -- both teams striving at full strength -- the final score is more indicative of the difference between the teams. And the stats would be consistent with that.

v-zero wrote:I like the idea, for the sake of a better game, however I don't like it from the point of view of statistical analysis. The game would cease to be (near) stochastic and all sorts of complexity in tactics would arise.

A hybrid format may contribute to some misleading statistical analysis, but I can't be sure if its effect would be more or less significant than the unusual style of play we see during the late stages of many games currently (with frequent fouling/foul shots, drastically shorter possessions, uncontested lay-ups, desperate three-point shots, foulouts, effect on margin of victory, etc.).

But "for the sake of a better game" says a lot - and may trump the statistical debate altogether.

I like the idea, for the reasons given. Given the inherent conservatism of fans and even more so the authorities in basketball, I think it has very little chance of becoming reality. But I think it is a good idea.

It took mens NCAA basketball, what, 30 years to adopt the shot clock? I became permanently turned off to mens college basketball because it was so boring for so many years. And even now the men and women both still have the idiotic 1-and-1 FT rule, which is an open invitation to foul instead of actually playing defense at the end of a game.

That 1-and-1 FT thing always surprises me when I watch the end of a college game. It seems to be a throwback to an era when FT were a primary way of scoring.

Of course, it's a large part of the strategy in the foul-athons late in games, because it significantly improves the chance that a team can erase a deficit in a short time.

It would be a fun and useful experiment to have games with 36 minutes of clock, and then play to 7+ points; and the 1+1 becomes redundant, so it can be dumped, in favor of 2 FTA.

I hadn't considered eliminating the 1-and-1 as part of a change in duration format, but I certainly wouldn't miss it if it were gone. I've never understood how the 1-and-1 contributes to a more fair game.

Every game should just be played to 100.
No game clock.
25 second shot clock.
When one team reaches 25 points, that is the end of the 1st quarter.
When one team reaches 50 points, that is halftime.
When one team reaches 75 points, that is the end of the 3rd quarter.
When one team reaches 100 points, that is the end of the game.

Beautiful.

Mike G wrote:That 1-and-1 FT thing always surprises me when I watch the end of a college game. It seems to be a throwback to an era when FT were a primary way of scoring.

Exactly. I don't know what FG% the college players were shooting back in say the 1940s, but BAA players shot 28% in the league's first season. So awarding a team a single FT was a reasonable reward (BAA players shot 64% at the FT line that season, again I don't know the college number). And when in the penalty, a 1-and-1 gave players significantly higher expected points than shooting FGs gave them.

But in the modern era, with FG%s significantly north of 40%, plus the possibility of a 3pt FG, a measly FT or 1-and-1 doesn't provide much of a deterrent to foulling, if a team is running out of time and is desperate to get the ball back.

In the better-shooting environment since the late 20th century, the awarding of two FTs is needed to properly penalize foulling teams. Actually I'd be fine with the NBA going back to the days of the 3-to-make-2 and 2-to-make-1, when in the penalty.

mtamada wrote:I like the idea, for the reasons given. Given the inherent conservatism of fans and even more so the authorities in basketball, I think it has very little chance of becoming reality. But I think it is a good idea.

It took mens NCAA basketball, what, 30 years to adopt the shot clock? I became permanently turned off to mens college basketball because it was so boring for so many years. And even now the men and women both still have the idiotic 1-and-1 FT rule, which is an open invitation to foul instead of actually playing defense at the end of a game.

Unfortunately, I think you're right. From what I can tell, it wouldn't be a lack of necessity OR soundness OR feasibility OR superiority (relative to the current format) standing in the way of a hybrid format - it would be a reluctance to change on the part of basketball's governing bodies.

Even a relatively minor game-clock-weaning change like using a points-based format for overtime only (ditch the game clock altogether during overtime and play first-to-ten or first-to-twelve or first-to-whatever) would seem to be too radical for the NBA or NCAA. (BTW I think this would be a major upgrade over the current time-based overtime, which is still subject to all of the game clock's negative effects.)

While I still think a hybrid format would be perfect for any league or level of play, it might be even MORE perfect for the WNBA, NBA D-League, etc. as a way to draw attention to the league while also improving style of play.