Would love to hear from anyone on other ways that I can determine the significance of these comebacks. I do not claim to be masters or gurus in statistics but will do my best to understand and learn from others on how scenarios can be measured. What I came up with could be different from what others come up with. And for that I would love to hear other ways with which I could determine "as accurately as statically possible" the zscores and how they matter for our the .
By using "weights", we are giving weights to each win and each loss based on the team's proportion of "wins" and "losses" within the total population (ie, the Majors, the National League or American League).
Simplifying that as much as we can. The statement "Mike batted .500 in 2019. Jim Batted .333 in 2019. Therefore Mike is a better hitter than Jim" is a case for weights. That statement is lacking sufficient information. Now lets say, Mike only had 2 at bats in 2019 and had 1 hit. Jim had 300 at bats and had 100 hits. With that, we would "weight it out" for a total of 101 hits (100 + 1)
Mike's weight per hit would be 1/101 = .010
Jim's weight per hit 100/101 = .99
And we would use those weights in place of the actual numbers so that the "proportions in relation to the sample will be accounted for. We don't claim for this to be the "end all and be all". It's just based on weight factors. We feel this gives a more "realistic" representation of "significance". If you know of other means that we could use that we are unaware of that can we could use, we would be very happy to hear from you.
The weights sort of help each team "co-exist" in the same population sample of "what it is we're comparing Team A with to Team B and Team C and so on..".
Thank you so much!
Baseball fan.
コメント