|
|
Post by volleydino on Feb 9, 2024 12:42:54 GMT -5
Reading the table, I see it has to do primarily with impact on a team’s offensive efficiency, given X person passing. A quality pass will translate more often to a side out. And the comparison here is within team - so by how much, given X passes, is the same team more likely to side out / score? I appreciate your take on it and agree with you that this is most likely the way that figure is calculated. I just totally disagree with it. The top passer is the guy who is passing the highest rating on a 4.0 or 3.0 scale. I don't think a top passer calculation should have any influence by a setter or a hitter, it should be on the actual pass. I get where you're coming from - it's a more cloudy measure, though it generally is correlated and important. In some way, I think especially for OHs and opposites, there is going to be a bias based on rotations (e.g. who is in the front row, when the OH is in the back row receiving?) Consider a more ball-control OH who has less attacking prowess. When that OH is in the front row, passing quality based on this measure may be biased downwards. (Unless he's tracking rotations and including dummy variables for that - in which case, that's great).
There seems like there would be less bias for liberos, since they would play all rotations in the back row. But their score is going to depend on who attacks off of the serve.
One interesting thing to note is that the middle attack is the most efficient attack. One factor that goes into whether a team is middle attacking is the quality of the pass. But I recognize there are lots of other factors too, namely the skill of the setter and the dominance of the middle attacker.
|
|
|
|
Post by jayz34 on Feb 9, 2024 13:18:46 GMT -5
I know he explained the rationale/methodology in the women's forums if you want to wade through his former posts. I think they are trying to factor in the strength of the server they are facing.
"Expected Expected Sideout" is how likely they think the passer's team is to sideout before the serve takes place based off how good the server they are facing is.
"Actual Expected Sideout" is how likely the think the passer's team is to sideout immediately after the pass (so it does not factor in what happens with the 2nd or 3rd contact).
|
|
|
|
Post by wilbur on Feb 9, 2024 13:23:12 GMT -5
I appreciate your take on it and agree with you that this is most likely the way that figure is calculated. I just totally disagree with it. The top passer is the guy who is passing the highest rating on a 4.0 or 3.0 scale. I don't think a top passer calculation should have any influence by a setter or a hitter, it should be on the actual pass. I get where you're coming from - it's a more cloudy measure, though it generally is correlated and important. In some way, I think especially for OHs and opposites, there is going to be a bias based on rotations (e.g. who is in the front row, when the OH is in the back row receiving?) Consider a more ball-control OH who has less attacking prowess. When that OH is in the front row, passing quality based on this measure may be biased downwards. (Unless he's tracking rotations and including dummy variables for that - in which case, that's great).
There seems like there would be less bias for liberos, since they would play all rotations in the back row. But their score is going to depend on who attacks off of the serve.
One interesting thing to note is that the middle attack is the most efficient attack. One factor that goes into whether a team is middle attacking is the quality of the pass. But I recognize there are lots of other factors too, namely the skill of the setter and the dominance of the middle attacker.
I would argue the most effective attacker in VB is not a MB but an OP or OH1 that is set when the middle could have been set and the MB makes a good pull and they are hitting against 1 or 0.5 pin blockers. More reinforcemnt about how important a good pass is. Looking from the other side of the coin, trying to score on a good mens college team that is passing well is really hard. Most digs, blocks and freeballs happen after a bad pass. |
|
|
|
Post by volleydino on Feb 9, 2024 13:36:07 GMT -5
I know he explained the rationale/methodology in the women's forums if you want to wade through his former posts. I think they are trying to factor in the strength of the server they are facing. "Expected Expected Sideout" is how likely they think the passer's team is to sideout before the serve takes place based off how good the server they are facing is. "Actual Expected Sideout" is how likely the think the passer's team is to sideout immediately after the pass (so it does not factor in what happens with the 2nd or 3rd contact). Ah, just found his GitHub and some relevant code. Not too complex at all. Something to go through when I have time. |
|
|
|
Post by chadgordon on Feb 9, 2024 16:01:02 GMT -5
Can someone please help me understand the chart. To me the biggest factor in determining best passer should be good pass % and passer rating. To me that should drop a few of the guys in the top 5 and should move some others in the 6-10 range up. Reading the table, I see it has to do primarily with impact on a team’s offensive efficiency, given X person passing. A quality pass will translate more often to a side out. And the comparison here is within team - so by how much, given X passes, is the same team more likely to side out / score? Yeah, basically looking at what kind of pass quality each individual server has created (so far this season), using the XY coordinates of where the next touch occurs, whether the MB was a threat, and factoring in aces/overpasses, etc. Once we know how strong each server is, we have an 'expectation' - then we look at the actual result from the reception and take the difference. You can look at this in terms of Sideout (where you're on a 0 to 1 scale... or 0% to 100%) or in terms of 'efficiency' and think about it more like attack eff and use a scale of -1.000 to 1.000. Some people prefer one or the other, but the gist of it is, how tough was the situation you faced (relative to an average team) and what was the quality of the reception (in terms of its value to an average team). You could argue that maybe teams with super strong MB would benefit more from a better pass than a team that only has 1 or 2 strong pins (where medium passes don't really change your sideout % much)... but tried to standardize across everyone, so it's obviously not perfect - but I can promise I'm trying my best to account for the context these guys are facing. |
|
