01-02-2013, 06:52 AM
I think Amber can be the kind of player we need, at least she has the physical talent. Previously, I had thought of Amber as showing more poor judgement than other players, which is often remedied by more experience. But what I saw in the UConn game was not simply passing into a crowd or not guarding her player well enough. She seemed to not be involved in the half court offense and not recognize she needs to hustle back on defense (neither of which shows up in A/TO ratio).
Remember, I only watched Amber in detail in a single half. Unfortunately, I don't have other games (this year) recorded where we played a challenging opponent. I should get those in the near future and can see whether what I saw is consistent or an aberration. (One can imagine that Amber had some physical condition that affected this one game, or she might have been coached to minimize her participation in the half-court offense in this game.)
Please remember that I stated that no one player lost the UConn game.
Assist to Turnover ratio has value but has its flaws. By making the ratio, you remove the playing time effect. (ie, I would argue that you want Assists/40min and TO/40min, but usually people just look at Assists and TO without looking at playing time. In essence, A/TO is a measure that compensates for whether the player played 10 minutes or 40.)Â
It seems a very good statistic for comparing two players with the same role on the same team who played in the same games, but you wouldn't want to compare a point guard to a post. It might have some value in comparing two guards on different teams if it was computed only for common opponents. But even then, there are problems with the comparison. One guard may play on a team that drives & shoots, while the other plays on a team of jump shooters. The turnovers they make may be similar, but the number of assists may be very different. Also, in general, if the a player played point guard on a high quality team, you would expect her to get more assists (because their teammates are better shooters or get into position better or are taller ...) than if she played on a low performing team. Using A/TO to compare players from teams that have few opponents in common is more questionable. Conceivably, you could tweak it to consider the strength of schedule, but such a measure would be quite inaccurate itself.
In Stanford's case, the assist totals were negligible in both the So. Carolina and UConn game. I don't think the cause in either could be attributed to the players who made passes, but the lack of assists does affect their A/TO ratios. In those games, the number of turnovers per 40minutes is a better way to judge the ball-handling performance at different positions.
So, if any player's A/TO is different this year than last isn't necessarily a good indicator that the player's ball-handling has changed. For Stanford in particular, I'd compute A/TO for the other games, and then consider the So. Carolina and UConn games as special cases.
Remember, I only watched Amber in detail in a single half. Unfortunately, I don't have other games (this year) recorded where we played a challenging opponent. I should get those in the near future and can see whether what I saw is consistent or an aberration. (One can imagine that Amber had some physical condition that affected this one game, or she might have been coached to minimize her participation in the half-court offense in this game.)
Please remember that I stated that no one player lost the UConn game.
Assist to Turnover ratio has value but has its flaws. By making the ratio, you remove the playing time effect. (ie, I would argue that you want Assists/40min and TO/40min, but usually people just look at Assists and TO without looking at playing time. In essence, A/TO is a measure that compensates for whether the player played 10 minutes or 40.)Â
It seems a very good statistic for comparing two players with the same role on the same team who played in the same games, but you wouldn't want to compare a point guard to a post. It might have some value in comparing two guards on different teams if it was computed only for common opponents. But even then, there are problems with the comparison. One guard may play on a team that drives & shoots, while the other plays on a team of jump shooters. The turnovers they make may be similar, but the number of assists may be very different. Also, in general, if the a player played point guard on a high quality team, you would expect her to get more assists (because their teammates are better shooters or get into position better or are taller ...) than if she played on a low performing team. Using A/TO to compare players from teams that have few opponents in common is more questionable. Conceivably, you could tweak it to consider the strength of schedule, but such a measure would be quite inaccurate itself.
In Stanford's case, the assist totals were negligible in both the So. Carolina and UConn game. I don't think the cause in either could be attributed to the players who made passes, but the lack of assists does affect their A/TO ratios. In those games, the number of turnovers per 40minutes is a better way to judge the ball-handling performance at different positions.
So, if any player's A/TO is different this year than last isn't necessarily a good indicator that the player's ball-handling has changed. For Stanford in particular, I'd compute A/TO for the other games, and then consider the So. Carolina and UConn games as special cases.