Planning youth soccer substitutions in advance for player development

I’ve always thought of myself as a pretty good youth sports assistant coach. Much like my professional life, I have never really strived to be the first in command. But I’ve always been more than willing to get involved however I can. I prefer the supporting role; in this case being a cheerleader on the sidelines for the kids and providing input for the head coach when needed.

I was an assistant coach in baseball for several years until my son fell rather deeply and quickly in love with soccer. It’s not a sport that was on my radar growing up, but as he started playing and following the sport I was quickly hooked. I helped coach in recreational soccer and futsal for a couple years before becoming an assistant for my son’s U10 club travel team for the 2016–17 season.

Me with my son Nolan Me with my son Nolan.

This past season (Fall 2017), I—for a variety of reasons—suddenly found myself in a head coaching role for the club’s U12 Boys team. The team I took on was something of a hybrid squad. There were four players returning from the previous year’s U12 squad. I brought five talented U10 players with me (including my son) who would step up to U12. Then I supplemented the roster with five recruits who were new to club soccer — some I knew and some were completely unknown.

I’m a very detail-oriented and analytical person. A huge reason baseball was always such a draw to me was the numbers. I’ve been into baseball statistical analysis since I was ten years old. Even when coaching my son’s baseball teams, I kept detailed statistics. I knew that if I was going to run a soccer team, I couldn’t do it like my U10 head coach did. Nothing against him—our minds just work differently. He’s a very laid back person who likes to substitute on the fly based on how the game is going. That’s something I knew I couldn’t do for several reasons:

First came the math — how could I arrange a 14-player squad into a 60-minute match where each player could get about the same amount of time on the pitch? It turns out this is a much easier problem to solve for a 13-man squad (which is what we ended up having in our first game) so I’ll show that first. Here’s how the 13-man setup works:

Position First Half Second Half
Min 0 10 20 0 10 20
GK P13 P7
LW P1 P5 P5 P10
ST P2 P6 P13 P4
RW P3 P8 P6 P2
AM P4 P7 P8 P3
DM P5 P9 P10 P9
LB P6 P10 P4 P12
CB P7 P11 P3 P1
RB P8 P12 P2 P11

The gist:

This worked well in our first game. We played well and even picked up a 4–1 win.

In our second game, we had all 14 players. The math here was a bit more difficult. This is what the lineup ended up looking like:

Position First Half Second Half
Min 0 15 22½ 0 15 22½
GK P14 P3
LW P1 P9 P2 P14 P7 P11
ST P2 P1 P12 P4 P5
RW P3 P7 P10 P2 P10
AM P4 P5 P8 P12
DM P5 P10 P4 P11 P6 P8
LB P6 P11 P8 P13 P9 P14 P9
CB P7 P12 P11 P2 P1 P7
RB P8 P13 P6 P13 P5 P13

The gist here:

We lost this game… by a lot. Pretty much everything went wrong, but one thing that stood out to me (and some of my players) was that we were substituting far too often. 7 or 8 minute shifts were not long enough to get a feel for the game. Just when a player was starting to get comfortable, I as asking him to come out or move. Not good.

The next time we had all 14 players present for a match, I’d streamlined things quite a bit:

Position First Half Second Half
Min 0 15 22½ 0 15 22½
GK P14 P5
LW P1 P9 P3 P13
ST P2 P7 P1 P2
RW P3 P10 P11 P7
AM P4 P5 P12 P3
DM P5 P11 P9 P11
LB P6 P12 P2 P10
CB P7 P13 P8 P4
RB P8 P6 P14 P6

Rather than having a ton of seven minute shifts, I tried to string three shifts in a row (more like 22 minutes) as often as I could. Previously, each position had 4–7 shifts over the course of the game with an average of 5.6 shifts. Now, each position had just four shifts each. The players responded to this system much better. In fact, they won the rest of their regular season games.

Going back to my obsession with baseball stats, planning my shifts allowed me to carefully track where each player was playing, for how long, and how often. If I made my substitutions at a time other than what my match plan said (usually only a minute or two difference based on when the subs were actually called in), I would simply note that change. After each game, I’d log the minutes in a spreadsheet.

Here are my players’ minutes by position for the season:

P1 15 26 20 130 20 114 7 260
P2 27 162 314 20 53 10
P3 27 80 15 127 43 236 2 55
P4 150 118 75 123 28 16 28
P5 7 265 138 61 7 37
P6 63 20 31 63 290 16 28
P7 25 10 28 169 182 58 30
P8 77 7 45 30 163 130 20 20
P9 92 20 374
P10 160 45 64 205
P11 56 10 34 26 49 104 56 100 30
P12 78 62 20 112 12 50 51 20 30
P13 20 10 47 10 25 28 117 120
P14 55 21 48 70 62 10 30

As you can see, I tried to get players time in as many positions as possible. P11 and P12 actually saw time at every single position while P3, P8, and P13 saw action at all but one. P9 only saw action in defense, but that was because of preference. I had a team of 14 players where at least 11 believed they were attackers. If a kid is willing and able to play defense and never wants to play offense, I’m going to oblige.

The benefits

When it was all said and done, there were quite a few benefits to this system.

  1. I went into each game knowing which positions the players would work on that day. While one of my goals was to win each game, I didn’t want to win at the cost of player development. This system ended up being a good way to make sure I didn’t do that. The only times I ever deviated from my match plan was when a player was forced to come off temporarily because of injury.
  2. The players knew what to expect. Before games, some of my players would ask where they were playing that day. But they knew that once a match plan was set, it was non-negotiable. At practices, I was happy to discuss where they would like to play in upcoming matches. As the season went on everyone became more comfortable with the system—even the players who weren’t asking me where they were playing each game.
  3. I avoided pigeonholing. I definitely had players who tended to be part of the attack much more often and had players who tended to defend. But I did my best to make sure those players still spent some time at other positions—particularly in practices and scrimmages. Did I do this perfectly? Certainly not. P10 didn’t play enough positions (I planned to use him in different places, but he ended up missing a couple games down the stretch), P6 (my son) was in defensive midfield way too much in the first half of the season, P2 probably should have spent more time as a midfielder, and so on. Tracking the player usage at least put these potential issues on my radar.
  4. I had the data. There were a couple times when parents or players would come to me with a concern. In most cases, I had data to show them that things weren’t happening quite the way they thought it was. But I still took note of their concerns and made sure they didn’t become an issue in the future, either. One example is a fun one—I overheard P2 complaining to last year’s coach that “Coach Adam doesn’t even play me at striker this year.” I responded with “I’ve played you at striker 112 minutes this year!” which elicited a chuckle from both player and coach.
  5. Everyone played. Remember that goal I had about every player playing at least half the time? I also tracked the percentage of possible minutes each player was on the field. Everybody played between 63% and 78% of the time. I’m proud of a couple things here—the fact that everyone played significantly more than half the time and that I didn’t lean on any players to play over 80% of the time.

Is this system for everyone? Maybe not. But it works for my risk-averse and analytical brain. I know I’ll continue to employ a variation of it in upcoming seasons.

Want to talk soccer? I have a separate Twitter account just for my footballing thoughts. Follow me at @fussballtwit.