Archive for November, 2007

First Attempt at Gathering Statistics

Friday, November 30th, 2007

So this week I took notes on my play, hoping to get an idea of what I should be paying attention to statistically.

I played first for my team through all three games, and played both ends of the alley in game one. Over three games, I played a total of 40 balls. Ten of those balls were leading off a round, 29 were lagging at an established point, and one was a raffa shot which went horribly wrong.

Game One:

  • I played 12 balls total
  • Leading off four times, I held the point against 0 balls (i.e. the opponents first ball took the point), 1 ball, 2 balls and held against all four opponents balls in the last frame. I think this is a pretty important stat and should be a good indicator of your lead-off lagging ability. If you can place the first ball close and make your opponents burn a bunch of balls chasing that you can gain a huge advantage. Further, you greatly increase your chances of throwng the last ball (or maybe two or even three balls after your opponent is done, which is huge.
  • The other eight balls here were all regular lagging attempts to make the point. Of these, I made point 7 times.
  • Our team scored 12 points, which puts me at 0.5 points per ball played. I think I need to sharpen this up a bit and start examining how many points I’m actually responsible for (either by directly taking the point or indirectly by either bumping a teammate’s ball into scoring position or by shooting an opponents ball that is beating a teammate’s ball, etc).
  • We gave up 2 points. Not bad, especially considering the team we were up against.

Game Two:

  • I played 14 balls, all from one end of the court.
  • Leading off four times, I held the point against 0, 4, 0, and 0 balls. Not so great.
  • Shooting once, I not only missed my target but hit my teammate’s (very well-placed) ball, costing our team two points. Disastrous. :(
  • Lagging against an established point nine times, I made the point four times.
  • On my end, we scored a whopping two points and gave up seven. We lost this game 9-12.

Game Three:

  • Again, fourteen balls, all from one end.
  • Leading off two times, I held against four balls and two balls. Much better.
  • Lagging in 12 times, I made point 9 times.
  • My end scored nine points and gave up two. We won this game 12-11 in a real nail-biter.

Now, this is pretty rough data. I’d be interested in something like lagging accuracy, some sort of average distance from the target, but getting that data would be cumbersome and nobody wants to wait around while we measure every single ball that gets played. Shooting accuracy is much easier since you either hit the target or you don’t. I’d also like to account for the fact that the distance to the target can vary wildly; I personally perfer to play a deep game, placing the pallino near the end of the alley when possible. This could have a negative effect on (e.g.) my teammates’ shooting accuracy. Someone on a team who plays the short game more often might have a higher shooting accuracy percentage but actually be an inferior shooter at equal distances. I think classifying shots as short, medium or long should be granular enough to make this sort of problem mostly go away.

I’m going to start keeping my personal stats in an excel spreadsheet, but I suspect that the data I collect in-game will change pretty significantly over the course of the season.

Bocce Statistics?

Friday, November 16th, 2007

I’ve been reading Curl with Math, which seems to have a very similar goal to this weblog. In curling, however, there are apparently a lot of compiled statistics available which make mathematical analysis a feasible activity. As far as I know, there are no such significant compilations of statistics for bocce.

So I’ve been thinking about this; if I were to start keeping statistics, what would I even want to record? Obviously, we’d want to start with number of balls played, how many points were made, how many opponent’s points were beat, how many shots connected. But this would only give an incredibly superficial picture of performance. Over the next few weeks I’m going to just record some info on all my shots and then see if we can come up with a bocce equivalent of RBIs or slugging percentage or QB rating.

Anyway, this week was actually quite uneventful (from an interesting-situation viewpoint). We dropped our first game 12-0, then won the next two 12-8 and 12-8. I didn’t really run into any particularly tough decisions during the match, but since we finished up early I got to watch a match between two pretty tough teams on the next alley. this situation came up:

14-Nov-2007 Image 1

14-Nov-2007 Figure 1

Red to play.

It’s hard to tell in the photo, but green is sitting on one point, but just a nudge to either his inside ball (Green B) or the red holding ball (A) will give him two points. The red team debated between lagging in along vector 1, trying to either beat the green A ball or to get lucky and maybe bounce off of the red A ball and sneak in behind for one point, or shooting hard along vector 2, hoping to pick the A ball off.

At first it appeared that lagging was the no-brainer. It seems much safer, but it’s a very difficult shot. Not only does the speed have to be just right to be successful, but the position of the pallino makes accuracy extremely important. There is a very narrow needle-eye that you have to thread in order to curve the ball in to the right spot. The chance to sell out and give green two points is there, but it’s very slim. The vast majority of times a lag is going to have zero effect either way.

Shooting, on the other hand, seemed suicidal at first glance. But a couple of things changed my mind. First of all, the two green balls are close enough that they make one effectively large target. It’s pretty easy to split them down the middle. If you make that shot, there’s a good chance both green balls will kick down the court and the played red D ball will basically die where it is. In general, a shot ball that hits a single target will carry quite a way, but one that hits two targets will usually transfer all of its momentum between the two. Further, the red A ball has the nail, so even if red knocks everything off the court, red still stands to make one or even two points.

The sell-out opportunities for shooting are really low probability. Basically your worst-case scenario is giving up three points, but doing so would require an incredibly bad shot. More realistically, red could pick off just the green B, having no effect, or kick green A into red A, which would probably give a net zero effect. The most important thing to consider here is that it’s relatively early in the game, so red shouldn’t mind a higher-variance shot here if it leads to a greater expected point outcome in the long run. Lagging, you’re looking at something like 1 point for green 90% of the time, and 1 for red 10% of the time. Shooting, let’s say you’re looking at 2 red 30% of the time, 1 red 20% of the time, 1 green 20%, and two green 30%. If that’s the case, your expectation for lagging is 0.8 green points, and your expectation for shooting is basically zero. You’ve got to take the shot.

Speaking of shooting, I took this clip of my uncle shooting with my cell phone. Next time I’ve got to remember not to stand in the ditch while recording. :)

(temporarily removed)

Week 1: Shooters vs. Laggers

Thursday, November 8th, 2007

One week into the season and I’ve already come across two excellent cases for examination.

Our league is organized into three-man teams. Last night, both of our teams were missing a man, leaving us with an interesting situation. The bad guys were two players who are basically pure laggers, while we had one shooter and one pure lagger.

The first things that should have popped into my head:

  • Shooting accuracy decreases much more dramatically with distance than lagging accuracy.
  • More balls in play favors shooters over laggers (assuming roughly equal players otherwise).

The reason more balls favors shooters is that your effective shooting percentage will probably be significantly lower than your effective lagging percentage; more balls means more chances to hit, and also means you’re gambling with a smaller percentage of your total balls for that round with each shot. On the other hand, if the players are significantly mismatched, more balls favors the better player since more balls equals lower variance per round. It’s easier to get lucky when you’re only throwning one ball than when you’re throwing three or four.

So, the first adjustment we should have made, but didn’t, would be to place the pallino closer when we’re leading off. I generally lead off for our team, and I like lagging deep precisely to defend against better shooters, but in this case I should have been playing a significantly shorter game. I think over the long term, this strategy could be worth as much as two or three points per game, especially if your opponents are super-reluctant to shoot at all. That’s pretty huge.

The next adjustment we could have made is really due to a loophole in our club rules. When a three-man team plays a two-man team, either all players play both ends, with the three-man team getting two balls per man and the two-man team getting three per man, or one player from the three-man team plays both ends while the other two play a single end each. The rules don’t address the situation where both teams are missing a man; sometimes both teams play three balls per man and sometimes they play two per man. We should have pushed for three per man (although I suspect we didn’t really lose much here since I doubt our opponents would have gone for it).

Now, for an actual in-game situation. This play came up at the end of game two. Green to play with one ball left. Red is out of balls. Green needs two points to win the game and is currently sitting on one point.

7-Nov-2007 Figure 1

Figure 1. Click for larger image.

Green is looking at two possibilities:

  1. Lag off the board and slide in for one more point.
  2. Shoot the pallino and take two points off the nail.

There are pros and cons to both of those. Most obvious is that if you shoot you have a chance of missing completely, or even worse picking off the green A ball, giving red two or three points. On the other hand, if you lag and come in just a bit too hot you can just as easily pop the pallino back into the red balls. The image doesn’t do a good job of showing this (I’ll hopefully be improving my diagramming as this blog progresses), but at the time the red D ball was in a position that made a lag here trickier than it looks in the diagram. It was not an obvious call either way. In the end, my assessment was that shooting would more frequently result in two points and a victory, but would also more frequently result in giving points away, while lagging was much safer, but much less likely to gain additional points. Given these choices, the only time you’d go for the shot and the immediate victory is if you feel that you’re seriously outclassed by your opponent. Taking an 11-8 lead is huge, even if you’re a slight underdog.