A poor attempt to explain the high-scoring S60-S63 era of the VHL

[studentized 166]

I had a hard time coming up with an idea for theme week this season. I usually like to incorporate a bit of data in my media spots, but the history of the VHL is already pretty well documented. Initially, I thought I would do some era-adjusted stat comparisons between current players vs those of the past, but collecting player stats prior to season 59 is still on my to-do list, and there wasn’t a whole lot of interesting player comparisons I could find with what I had. It wasn’t all a waste however, since all that work did eventually lead me to something I did want to write about, albeit with not a lot of time left to actually do the writing.

 

The chart below plots the scoring output for how the league-average player performed across each season from S59 to S68. It may not look like there is a huge difference (in hindsight, comparing the “league-average player” across seasons may not be the best way to view what I want to show), but the effects become more obvious when looking at some of the stat leaders; S62 was by far the craziest.

 

 

 

I was already somewhat aware about the high scoring S60-S63 seasons, just from chatter around the forum, but what I never really understood was what caused it. Why did scoring come back down to earth in S64 and plateau? What caused it to go up in the first place? Naively I just assumed that the SHTS sim has lots of knobs that can be tweaked and that some knob was probably turned in such a way that made scoring increase. While possible, that answer is highly unsatisfying. More realistically, there is probably something in the data that can explain it. I had a couple of ideas about what types of things I would think would cause a scoring increase in the VHL.

 

For example, maybe the ratio of the number of user-created forwards to the number of defencemen/goalies was more in favour of forwards? If there weren’t a lot of people that chose to make players in goal-preventing positions, then it would make sense to see scoring go up. The chart below plots this information. As you can see, this mix of positions has pretty much been left unchanged across the years. In particular nothing here stands out about S62 that would explain it's league scoring rates.

 

Rather than looking at the quantity of players at each position, we should really be looking at the quality. It still feels like TPA is the most common way that people measure the quality of a player on the forums, so that’s what I used. Below plots the average TPA invested in each player at each position group (forward, defence, or goalie). S62 clearly looks like the outlier here, with low average TPA across all position groups. Most notably it was the worst goalie crop we’ve seen in a while, mostly due to retirement of the elite guys from the year before. But in terms of the gap between TPA in forwards vs TPA in defence vs TPA in goalie, there really doesn’t seem to be a clear explanation. In fact, the TPA gap favours forwards in more in recent years than at any other point in my data, yet league scoring hasn’t returned to it’s past high.

 

 

                                                                  Number of User Players By Position                                                           Average TPA of Players By Position

 

One obvious flaw to the above line of reasoning is that it doesn’t account for where the TPA was spent. There are definitely high TPA forwards that don’t score much (Current Riga forward Anthony Matthews has the highest TPA “cost” per goal scored in all my data). And there are also high TPA defensemen that don’t prevent scoring well (this is harder to measure, obviously, but the most notable high TPA, high minus player from my data was Luc-Pierre Lespineau-Lebrunette. What a fighter). To account for this without getting too complicated we might only want to look at two TPE stat categories; SC as the strongest predictor of goal scoring and DF as the (most-likely) strongest predictor of suppressing goal-scoring. Below the chart plots the average rating of SC vs DF across seasons. One thing this shows is how investment into DF changed after S63. Despite SC increasing on pace with DF in recent years, it seems that just the presence of DF being invested in played a role in moving past the high scoring early 60’s.

 

 

In my mind, there’s still a lot left unexplained in regards to league-wide scoring variations across seasons in the sim. What I am more confident in saying is that the early S60’s were likely an anomaly; a period where the average player was worse than we’ve seen in a long time, where people didn’t care as much about DF as we currently do, and where the few good players that did exist were able to take advantage and score at an insane rate. I doubt we see something quite like it again.

[Gustav 6,407]

The S63 draft was also fairly large, so the league saw a good group of low-TPE guys entering the VHL that season. I'd be willing to bet this evened out depth a bit and made it so teams weren't just the same three elite guys repeatedly putting the puck in the net.

 

Since then, too, draft classes have been good-sized--the S66 class was historically big, and we've seen classes at or near that level since.

[Tagger 2,722]

Nice article! As someone who was there, I have a couple of reasons I feel that could have contributed to the windfall of goals. 

 

1) Team setups - It wouldn't surprise me if some of the teams in S62 had the largest SC:PA ratio in the VHL history. Lots of the players on the rebuilding SEA and RIG teams were sitting on 40 passing, but still had 90+ scoring, and this effect on the decision making formula likely, combined with the DEF drop you mentioned, led to the huge amount of shots. 

 

2) Goalies -   The goalies took a real hit in S62 in comparison to S61. In Season 61, the average overall goalie was 73.9 Overall, and this went WAY down to 70.1 Overall in Season 62 before going back up to 73.4 Overall in Season 63. The dip in goalie talent combined with the greater number of shots spelled disaster for a lot of teams. 

Edited October 21, 2019 by Tagger