Everything You Ever Wanted To Know About Serve Zones (Part 1)
Taking questions from VolleyTalk really seriously...
Note: This post got so long I had to split it into 2 parts as it went over the Substack email length limits. Here’s Part 2.
A poster in a recent Volleytalk thread asked:
I remember seeing a stat diagram about SO% based on where the server stands behind the end line. I remember down the line being the lowest S/O% for the receiving team and a player serving out of the middle was super high SO%. This may have been a Joe Trinsey or Kessel article possibly, but I'm not 100%.
Does anyone have any clue what I'm talking about AND where to find it?
When somebody throws up the Volleytalk bat signal, I have no choice but to respond.
There’s a few related topics here:
Where is the best place for the server to serve from? (We’ll call this “Start Zone”)
Where is the best place for the server to serve to? (We’ll call this “End Zone”)
What are the best combinations of Start Zone and End Zone?
How is this affected by which passer receives the ball?
How is this affected by whether said passer is frontrow or backrow?
I’m going to keep this post free but, at the risk of sounding super-needy, please consider subscribing if you find this info valuable. Some of this information is the distillation of hundreds of hours of study over the past decade or so, both by myself and other people. Special hat tips to Nate Ngo, Jeff Liu, and Peter Wong; as well as other coachings I’ve collaborated on various projects with. At this point I sometimes lose track of who I did some of this research with, bear with me…
Serve Effectiveness By Start Zone
Okay, so don’t serve from Zone 6. Got it. But also: nobody serves from Zone 6.
Now you have the slight issue about what’s considered to be “Zone 5” or “Zone 1” and what’s considered to be “5/6 Seam” or “1/6 Seam.” In Volleymetrics they are coded (5/6 Seam as “Zone 7” and 1/6 Seam as “Zone 9”) and the coding is reasonably consistent, but…
In my opinion, “Zone 5” and “5/6 Seam” can be effectively batched together. “Zone 1” and “1/6 Seam” can be effectively batched together. I think for practical purposes, there just isn’t much separating them there. I think you’re losing some nuanced analysis of what servers who hug the sideline produce vs off-sideline servers, but there’s just no way to mass-analyze the data without pulling in VM data, and this is how it’s coded.
So let’s say that the “Start Zone” chart looks more like this:
So:
Serving from Zone 6 sucks.
Nobody serves from Zone 6 anyway.
How about looking at some non-Volleymetrics data
This was from a previous study that I did, and I coded the start zones by hand. I believe this is a more true differentiation of the 5 potential start zones. It does seem to show a trend that fewer servers are serving out of true Zone 6 than previously.
I stupidly didn’t record Sideout % off these serves, I only looked at KO (Knockout) %, a measure of whether you (a) served the ball in and (b) knocked the other team out-of-system. And I now only have the results of this analysis, not the original codes themselves. So, take it for what you will, but it also does seem to show some trend of servers being more effective when initiating their serves closer to the sidelines than the center of the court.
Okay, so that’s Women’s NCAA, are there any other levels we might would look at? In Part 2, I’ll have some men’s data on this.
Serve Effectiveness By End Zone
In a previous post, I note that it’s not a good idea to serve Zone 6.
I don’t think that’s breaking any revolutionary ground.
That’s also a really strong trend. 2% difference in sideout is worth 2 or 3 full match wins over the course of an NCAA season. So if you could snap your fingers and have your servers put all their serves in 1 and 5, you would instantly be a better team. Of course, it’s not that simple, there’s other factors involved. In the article, I make a case for serving straight-on, or at the very least, having the ability to do so.
Let’s look at some more data.
This is hand-coded by me and it includes some more nuanced data. I’m using the seam terminology we used at USA, where we highlighted the 4 seams that are open in a traditional 3-passer alignment. In this study, I excluded serves against hard pinches or 2-person reception patterns, which would have different seams. But since a 3-person reception pattern is by far the most common in higher-level volleyball, there’s some actionable data here.
The first big takeaway is that simply serving the ball into a seam is a huge win. Passers in any zone were fairly competent when the ball was served at them, and struggled when the ball was hit to a seam. There’s two other clear takeaways:
Passers struggle more with the sideline seams (44% and 50% Good Pass in the 1 and 4-seams compared with 57% and 56% Good Pass in the 2 and 3-seams) but…
Some of the advantage of serving the sidelines is nullified by increased errors.
When doing this study, I included service errors into the seam that the player appeared to be attempting to serve. So, if you missed long in the 2-seam, I assume you were aiming there. If you missed out the zone 1 sideline, I assume you were aiming for the 1-seam. No method is perfect, but I think that was a reasonable way to assess.
What Are The Best Start And End Zone Combinations?
That study that I did in 2016 was compelling off for me to start deliberately training servers to serve from the sidelines to the sideline seams. Translating any strategy from theory to practice is messy and there’s no guarantee. Anecdotally, I’ve liked the results. Anecdotally, I’ve also seen servers struggle to hit the sidelines and been incredibly frustrated with multiple sideline errors when we could just take the sideline out of play and aim for the interior seams.
Another anecdote, but in reviewing our recent VNL for Canada MNT, we gave up more than twice as many interior-seam aces than sideline-seam aces. This tells me that even the best servers in the world may not be able to pepper the sidelines all the time. However, I think there’s a bit of a twist there when you look at float servers vs spin servers. Max-power spin servers are naturally going to aim a bit more into the court to reduce their margin of error. High-accuracy float servers should (and in my experience, are) able to zero in on a more specific target with repeatability.
But let’s stick to float serving for now.
So there’s a pretty compelling case to serve from the 5 sideline down the line to Zone 1. And you’d also note that most of the time this would be done against a non-libero in Zone 1, particularly in Rotation 1. Again, you see the devastating effect of putting the ball on the Zone 1 sideline, and again you see that effectiveness about halved by missed serves, relative to serving the interior seams. The lesson here might be that accurate servers should try to hit the sideline while less-skilled servers might want to serve the interior seams.
What’s also interesting is that the Zone 6 passer was actually pretty bad at passing this ball as well. 51% Good Pass is pretty low at this level. When I studied seam relationships at USA, I found a similar pattern in that seam.
Serving Zone 1 in general is a common serving strategy, particularly when it’s a weaker-passing outside. But what’s interesting is that, across the board, Zone 6 passers also seem to struggle with passing this ball and probably over-poach it, at least at the FIVB level.
What about the other side?
A similar pattern, although hitting either one of the crosscourt seams also seems to work. But that sideline seam from Zone 1 is even more devastating. And there’s a logic there (that we’ll explore in the next post), you’re more likely to have a weaker-passing outside in Zone 5 than Zone 1. The libero never passes in Zone 5 in the traditional reception alignments.
What’s interesting here on the 1-to-5 serve is that the cost of missing into the interior is significantly higher than with the 5-to-1 serve. And this makes sense. At the risk of repeating “in a future post” again (but seriously, I’m running up against Substack email length limits), we’ll see that Zone 6 passers are significantly better on their left than on their right, which I have to imagine accounts for the fact that hitting the 3-seam from 1 is not nearly as good as hitting the 2-seam from 5.
A final graphic to wrap this up with:
We see that, combining everything from this whole article, there’s nothing worse you can do than serve from the center of the court to the center of the court. That’s a fact. It’s my opinion that the easiest way to combat this is to, generally, get your server closer to one of the sidelines. And keep the ball out of the center of the court by putting the ball down the line.
There’s a whole host of exceptions and tactics and plays and counterplays. But I’d start with just getting your servers over near the sidelines and working those seams, and seeing if any of them can reliably paint the sidelines down the line. With how effective those serves are, everything else is gravy.
Thanks for sharing! Super interesting.