Apologies: In an earlier post (#12427), I made a mistake in computing how many games play throughout their career based on draft age.
Here's the correct numbers (effective games = total minutes/48):
Players drafted in calendar year they turned 18: 144
19: 213
20: 240
21: 169
22: 145
23: 122
24: 90
25: 20
But because this can be greatly skewed based on where a player was drafted (older players generally drafted lower, therefore it makes sense that they play total fewer "games."), I came up with another measure to show the risk/reward scenario for drafting players at certain ages. This is based on exceeding the 75th percentile of all draftees in expected total win shares, or failing to reach the 25th percentile:
18 year olds: 40% expected about 75th percentile (based on draft position), 40% reached that level; 30% expected to be below 75th percentile, 50% were below.
19: 31% expected, 42% actual above 75th; 8% expected, 42% actual below 25th
20: 40% expected, 45% actual above 75th; 13% expected, 36% actual below 25th
21: 25% expected, 29% actual above 75th; 21% expected, 54% actual below 25th
22: 26% expected, 27% actual above 75th; 27% expected, 60% actual below 25th
23: 17% expected, 17% actual above 75th; 30% expected, 67% actual below 25th
24: 20% expected, 10% actual above 75th; 45% expected, 80% actual below 25th
So no real surprises compared to conventional wisdom. This goes along pretty well with what I wrote in the earlier post. If you want someone to exceed their draft position, you're somewhat more likely to find that with a 19 or 20 year old. 19 year olds are more risky, but carry somewhat higher likelihood of a higher reward.