Everyone’s favorite!
I’m building out a stat summary dashboard and I want to display team power play conversions.
Let’s say we have the following sequence:
- Player #2 on Team A is excluded
- 10 seconds into the power play, Player #3 on Team A is excluded
- Player #2 returns to play
- Team B scores before Player #3 returns
If this was the only exclusion in the game, would you say that Team B’s conversion rate was 50% or 100%?
Interesting question- I would say 50%.
Personally I count it as a successful 6 on 5 conversion, not impacting that statistic, but also keep track of other man-up opportunities and would mark us down as 0/1 on 6 on 4 even if it was only for :10.
100%. It was a single man up situation.
I would call that 100%. To me, it was scored on one extra player possession that started when player #2 was excluded, was extended by the exclusion to player #3 and isn’t over until #3 returns to play.
Edit: And I think to make the argument make sense make the example more extreme. If on a single possession one team had all 6 field players excluded and then the other team scored and no other exclusions were called the rest of the game, it wouldn’t make any sense to say they went 1/6 (16%) and I don’t think anyone would argue otherwise. And the reason it doesn’t make sense is that the exclusions all occurred on the same possession. If the exclusions overlap, all any subsequent exclusions after the first do is extend the amount of time the offense is playing an extra player possession.
I would count it as Gibson and Breck. I think it has to be that way. In addition to Breck’s points, I would add that if the goal in your scenario had been scored a bit earlier while both #2 and #3 were excluded, you wouldn’t count that as two power play goals being scored. If you did, then it would throw off the calculation for how many of a team’s goals were power play versus natural.
To keep all the stats making sense, a power play has to be viewed on the team level. So long as a team is man-down, the opponent is in the power play. How many men-down or how long the man-down condition persists, including between periods, is irrelevant.
In your hypothetical, there was one power play and one goal scored. If that was the only goal the team scored, then 100% of the team’s goals were power play goals, and the team converted on 100% of their power play opportunities.
I think y’all have talked me into the conversion rate being 100%.
I should have predicted that as it’s the more difficult one to code 
Thanks everyone!
Just to continue the talk…I love talks about stats and water polo…do you all think that breaking things dow like 6 on 5 v 6 on 4 is too much?
Probably something a coach would have a stat keeper do like tracking shots made/missed against their goalie and where they were taken in the pool.
I have a similar question to this. How are goals attributed when scored during the 4 minutes of power play after a major misconduct for violent action?
I ask specifically for calculating things like a conversion percentage. Does it mean that every possession during that time is considered a new extra player possession?
As with the above, I don’t think there is a “right” answer as much as there needs to be a consensus and standard held.
My philosophy has always been that if you have the ability to track something you should do so.
Sure, 6 on 4 doesn’t happen all that often, but in tracking the scoring etc for those situations both the offensive and the defensive team could learn something about how they are having success or struggling and the coaching staff can then work to address that.
For instance, I’ve found that a lot of teams struggle in the transition from the 6 on 4 back to a 6 on 5 and I think you might be able to see this with stats about when the goals are being scored (as in the original example that started the whole thread) and target that for improvement for your team.
Hi,
No. Doing the statistics, you focus on natural goals vs. man-up situations. 6v5 are goals that should be tracked as they have an impact on overall shooting percentage, teams make set plays and can go back and determine if the set-play worked, needs refinement or scrapped all together.
You will hear listening to announcers the impact of a player making a ‘natural’ goal vs. a power play. A natural goal is of course a goal that is scored without a penalty shot. This is important to distinguish when reviewing the game if a player is 3 for 7 (.428), but 2 of the 3 are penalty or man up situations, then you have an issue of a player actually 1 for 5 (.200), which changes the after-action review of the game. Was the player taking a unbalanced shot? was it hurried? Was the player out of position? The stats and AAR allows to deep dive into this information.
I think your point about a power play leading into a penalty being registered as a successful conversion of the power play regardless of the result of the penalty is interesting and I think I agree with the thinking behind it.
It is similar to how I’ve always felt the player who draws a penalty should be rewarded somehow in the stats with a half an assist or something on a successful conversion.
This brings up another issue I’ve had with the official stats provided by WA etc. They make no distinction between drawing an exclusion and drawing a penalty, but those two opportunities are not created equal. I believe exclusions drawn and penalties drawn should be separate stats.
Taking NLM one step further, and it would be subjective, but quality of shot should play into it. If the shot is an end of clock lob while the rest of his/her team is already back on defense there should be no negative attributed. An early shot in the clock from 8 meters with a defender in their face should absolutely get some negative points attributed.
Likewise, a missed shot that the goalie tips out is not the same thing as a shot directly into the goalie’s stomach.
This is where something like expected goals (xG) from soccer would be really interesting and useful for water polo. It adds all that context like shot location on cage, location of the shooter, location of the defenders, etc beyond just whether a shot was taken and it missed or went in.
Such an xG model could feasibly put together using the shot charts etc that apps like CB and Total provide, but you’d need access to that data and you’d need a lot of games in the database to build the xG model.
From a few years ago, my business partner Joey Gullikson (UoP asst. coach circa 2010s) put out a few write ups for our blog a few years ago:
My dataset has some muddy information on shot location, target on goal, and situation (even vs power play) but not as rich as what they used in the study the article references.
I wonder if a “good enough” xG could be derived from this notion.
I like the framework of assigning values to stats that represent a “subgoal” but I’m not sure if I agree that a missed shot should be negatively attributed.
While you are giving the opponent the ability to score, I have some poorly articulated notion that the potential upside of a missed shot balances out with the downside of giving the opponent a possession (something-something “can’t win if you don’t play”).
Disclaimer, I haven’t dug deeply into the underlying data in a while BUT regarding symmetry on outcome depending on side, I’d offer a different hypothesis:
The sample was NCAA level so I’d guess that shot selection skews the data. That is, a right-hander shooting from the “left-hander’s side” understands they’re in a disadvantaged position and thus select for shots that they understand are likely to score.
I also don’t know if the sample size is comparable from each position either.
I’m very curious to see if there’s still symmetry when we’re shooting in a controlled drill / experimental setup vs live game observations.
which questions the emphasis put on handedness.
Indeed, I remember a chat (on this board somewhere?) where the topic of USA WP not rostering as many lefties was discussed. Makes me think the performance coaching staff is way ahead of the curve 
!00% on 6on5, 0% on 6on4.
That’s a lot to process.
Thinking about this more and trying to judge what is a good shot vs. a bad shot.
Probably more specific to age group level instead of college, but a good shot for a team’s “shooter” from 5 or 6 meters is a bad shot for a couple other kids in the pool. If my shooter is open or has a clean lane from 6 or 7 meters I will take that chance most of the time and if I’m scoring it. I would not assess a negative value even if he/she misses the shot. That same shot from less developed players/arms is probably a shot I would not like to see and might assess a negative value to that shot.
In order to correctly score a game in this manner it takes someone with knowledge of the players and a lot of time after the game to go back to the game film.
Another variable to consider is the goalie. That same shot from 6 or 7 meters against Weinberg is not the same as against an average or below average high school year old goalie. That’s extreme, but for the sake of argument, we could say a national/ODP level goalie vs. an average goalie. I would start assigning more negative values for shots that we know are going to be saved by the superior goalie while we would be asking for more outside shots against a subpar goalie.
So could a scoring system adjust from game to game?
