Posted by Benis on 10/30/2019 3:08:00 PM (view original):
Posted by mlitney on 10/30/2019 1:54:00 PM (view original):
"5) Related to 4, the final odds you see have been stretched to favor the winner. So the “underdog” is quite a bit closer to you in terms of accumulated effort credit than the odds show. If the odds weren’t stretched, that 74-26 battle would have been more like 61-39. It’s still an upset and a surprise, which is in the game by design because the developers want some surprise and unpredictability; but without the stretching, while it wouldn’t *feel* like such an upset, it would mean the leader would lose more often."
Ive never heard of this "stretching" before. Is that something from seble, or is it from some mathematical calculation you've done over the years? I'm curious to understand the need for this, and how it impacts recruiting. It shows teams being farther apart than they actually are? Or does it actually give an advantage to whichever team is the leader?
Seble described this in Beta. It only adds a few % points. Not as much as your example here.
False. It’s a sliding scale, so the difference is small when it’s very close - for example, 50.1 to 49.9 will be “stretched” to 51-49 and 54-46 might be “stretched” to 58-42; but the boost gets larger as the effort credit difference grows.
We know this is true, because seble explicitly told us at one point that 37 in a 63-37 battle was not in signing range. Close, but no cigar (I’m paraphrasing). So the signing range cusp is somewhere north of 37 in a 63-37 battle. Or another way to look at that is to get roughly 60% of the effort credit the leader has accumulated. That gets you *barely* into signing range, at something like 80-20, which is the longest odds I’ve heard of.