Friday, December 11, 2009

Movements vs. Movements

Today is the 1st anniversary of the 1st post to the BM Report. There have been many days when I have considered retiring from this endeavor. It seems like a lot of work for little payoff, but four people seem to enjoy it (on a good day). So, I'll keep slogging and blogging. Anyway, given the occasion, it seemed like a perfect time for some hindsight and reflexivity. So that's what we'll do.

There is no greater opponent than yourself. If you had to battle yourself in a fight to the death, who would be victorious? You would, of course, which is why when we fight ourselves, one of us must have a name preceded by the title "Bizarro". What if you could battle yourself from last year? Would the younger, slimmer version win or the older wiser one? This is today's question. If the Movements of last season rolled against the Movements of this season, who would be victorious?

To answer this question, I went into the closet and dusted off the old BM Simulator. This funky little tool simulates games of bowling. To do so, it needs two pieces of information: 1) The likelihood of any given 1st ball score (zero to strike); 2) The likelihood of picking up a spare depending upon how many pins are left standing. These probabilities are easy to estimate if you have enough data.


Here's how the Movements of olde tymes stack up against the 09-10 BM's. Not surprisingly, we are very similar teams. The graph above shows the probability of getting any given score on the first ball. Where it really matters, 9's and X's, this year's team comes out on top, but only slightly. Interestingly, this team is also slightly more likely to chuck the 1st ball into the gutter.


Things are again very similar for pickup %. This year's team has improved our single pin rate, but struggled a bit more with 2 and 3 pin leaves, no doubt due to the proliferation of splits. The biggest difference between this year and last concerns 8 pin leaves, or a 1st ball score of 2. Last year, we converted 75% of our chances. This year, we haven't done it a single time. This difference, however, is not real because it happens only very rarely. From the prior graph, it is clear that 8 pin leaves are the least common type. We only had four all of last year, and this year, we are 0 for 5.

So when these two teams are pitted against one another head to head, who would be victorious? To answer this question, I simulated 5,000 games for each team. I then put them side by side and counted how many games went to the new or the old team, and how many games ended in ties.

The good news is that this year's team would beat last year's in a battle to the death. It is not a big difference, but a difference it is. The 09-10 BM's would win about 52% of the games. Those sorry losers from last year would get 47%, and 1% of games would end in a tie. In the simulated games, last year's team averaged 144, and this year's 146, a result that is remarkably similar to real life. We currently average 147 as a team (including K-Terk).

So, this is a long and convoluted way of saying that over the 1st year of the BM Report, we have gotten better, not a lot better, but improvement is a good thing. If we meet those chumps from last year in a dark alley, and must fight to the death, I will be confident, 52% confident.

4 comments:

  1. Happy 1st Blogoversary, brother!

    ReplyDelete
  2. Thanks, E.B. It seems like it has been longer than a year.

    ReplyDelete
  3. What's the win percentage without me included? I have to have brought this year's average down a bit. I've also raised the team's good looks, charm, and sense of humility a great deal. I'm so humble, it makes me sick!

    Congratulations on the anniversary- without this blog, I be much more productive at work in the morning.

    ReplyDelete
  4. Well, we're 4-8 with you and 20-20 without you. Now, I know who to blame. Get back to work, k-terk.

    ReplyDelete

Note: Only a member of this blog may post a comment.