I missed this earlier this year, but Musings Marcels held their own against other, more complicated prediction systems. I found this paragraph most interesting, however:
Averaging the projections might be a great idea. The “Average” row in the above table is exactly what you would expect: the accuracy of the average of all four systems. It beats all four systems in four of the five categories, and fell short of only Steamer in the fifth. One would expect that an average would rarely be egregiously wrong; it’s surprising to see that the average also tended to be closer to right than each individual projection. This could be a quirk of a single season of projections, but at the very least, it seems to say that the brute-force method of resolving differences between the projection systems is credible.
I worked for a few years in information retrieval. What was becoming clear at that time was the aggregation of multiple search engine results did better than any one individual search engine. The weaknesses tended to wash out, so I’m not at all surprised that an average of the four projection systems does better than any one individually. I suspect this will work for WAR as well. Take three different versions of WAR, average them together, and you probably get a more accurate view of a player’s value that any with any single WAR measure.
from baseballmusings.com http://ift.tt/2mxTHWR
No comments:
Post a Comment