Confounding Factors in Bowling

Just stumbled across this article, The Hardest Shot in Bowling. It’s a really interesting read, and for someone like me (who loves to play with data of all kinds!) it is always great to hear of new data sources.

However the article did make me wonder about the accuracy of predictions. It seems (though we can’t know 100 %, as the methods weren’t fully discussed) that Blatt has used a simple probabilistic model to determine the “Hardest Shot in Bowling”. And indeed, it is hard to argue with the cold hard numbers presented. But perhaps there is more going on here that we think…

I wonder how likely it is that a given player, P(i), will face a Greek Church formation? I wonder too, if we had some measure to rank players and the likelihood of facing an unfavourable split, whether we would see a slightly different story appear?

For example, let’s say there are two friends, Tony and Jim, who play every week. Tony is arguably a much better player than Jim and he never faces a Greek Church split. Now Jim isn’t so lucky, over the course of a year he throws down 10 tricky balls that result in a Greek Church split, of which he only manages to convert 2 out of 10.
false_probs
Based on the raw numbers, you might say that there is only a 20 % chance of a spare when faced with a Greek Church split. But of course that isn’t the whole story – because the player matters! If we consider the player in our model, then Jim has a 20 % chance of converting a Greek Church split. But we don’t have any data about the odds that Tony might convert this formation.

Now, I have no beef with Blatt’s analysis. It is truly interesting. But it just seemed like a great example of potential bias. We always want to simplify our models and solutions; to paraphrase Einstein, things should be a simple as possible and no simpler. However there is the risk that when we over simplify the numbers, the story can be quite misleading…

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: