OOTP Developments Forums - View Single Post - 20-80 ratings question

NoOne · 02-19-2018, 04:53 PM

that's probably certain -- the 20 vs 80 comparison... the closer they 2 numbers are the tougher it is to say wiht such confidence.

also, your current ability might be adjusted by settings... not limited to AI evalution using stats and ratings in a calculation of current ability.

if the difference isn't so large, it may or may not be true, though... obviously the larger the difference is, the more likely the higher value is 'better'.

why: it may not scale to statistical environment (still holds even if it does). different combinations of ratings mean different things, even though they add up to the same current/potential value.

higher potential correlates to "better" is the best way to say it. no causality there on its own. it's a calculation of many things that are individually cuasal in nature.

potential is just untapped ability... if they are still young you can assume or still hope that they'll develop, and if they are age 27-28ish you can assume it will be left untapped most of the time. take them as they are, not for what they will become at ~28... you draft for what they will become, hopefully become that is.

it adds up the same way, just using potentials for every rating in stead of current ability.

with 100% accuracy most defintiely you want to pay attention to potential... other factor is that an 18y.o. is more likely to gain velocity than a college aged kid. so, their stuff, even potential stuff, may be underrated.

i'd break it into tiers -- your choice(s) on the range(s). if you are using the "by 5" 20-80 scale that's only a "13-pt scale", and you should use larger ranges with lower resolution scales.

when players fall into tiers i then look at their individual ratings. if you don't have 100% accuracy, you want to choose based on traits within that "tier" that's suitable for that round/pick, not potential.

even with 100%, i'd still pick based on individual ratings of that player, but it would inevitably have to be closer difference between potientials. ("potential" is more stongly correlated with success @ 100% accuracy, doesn't mean 80>79 potential, though, just more likely).

02-19-2018, 04:53 PM	#2
NoOne Banned Join Date: Apr 2015 Posts: 7,273 Infractions: 0/1 (3)	that's probably certain -- the 20 vs 80 comparison... the closer they 2 numbers are the tougher it is to say wiht such confidence. also, your current ability might be adjusted by settings... not limited to AI evalution using stats and ratings in a calculation of current ability. if the difference isn't so large, it may or may not be true, though... obviously the larger the difference is, the more likely the higher value is 'better'. why: it may not scale to statistical environment (still holds even if it does). different combinations of ratings mean different things, even though they add up to the same current/potential value. higher potential correlates to "better" is the best way to say it. no causality there on its own. it's a calculation of many things that are individually cuasal in nature. potential is just untapped ability... if they are still young you can assume or still hope that they'll develop, and if they are age 27-28ish you can assume it will be left untapped most of the time. take them as they are, not for what they will become at ~28... you draft for what they will become, hopefully become that is. it adds up the same way, just using potentials for every rating in stead of current ability. with 100% accuracy most defintiely you want to pay attention to potential... other factor is that an 18y.o. is more likely to gain velocity than a college aged kid. so, their stuff, even potential stuff, may be underrated. i'd break it into tiers -- your choice(s) on the range(s). if you are using the "by 5" 20-80 scale that's only a "13-pt scale", and you should use larger ranges with lower resolution scales. when players fall into tiers i then look at their individual ratings. if you don't have 100% accuracy, you want to choose based on traits within that "tier" that's suitable for that round/pick, not potential. even with 100%, i'd still pick based on individual ratings of that player, but it would inevitably have to be closer difference between potientials. ("potential" is more stongly correlated with success @ 100% accuracy, doesn't mean 80>79 potential, though, just more likely).