Login

**RomanesEuntDomus** · 01-10-2020, 11:50 AM

Hello SHL. Over the past few weeks and months, there have been a lot of discussions surrounding the newly discovered Passing-gap exploit, which has only picked up steam recently given Hamilton’s dominant performance in these playoffs using that very exploit. However, I would argue that the conversation about this topic has become very flawed and detached from the facts at this point, so I decided to write an article about this to summarize some of my criticisms. To my own surprise, this article ended up being more or less a Defense of the Hamilton Steelhawks and therefore it will be called:

Making the Case for the Hamilton Steelhawks

Be warned however, this is only the first half of a two-parter, as my next article will be about “Making the Case AGAINST the Hamilton Steelhawks”. I had planned to actually do the two together as one long article but ran out of time, so I’m putting this part out first. Expect the second one to be a bit shorter though both because I won’t have that much time, and because I would argue that the arguments against the Steelhawks are already way more prominently known with people vocalizing most of them frequently, so I don’t see as much of need of making all of them in all detail again. Instead, I prefer to play devils advocate here. So let’s begin…

1. It’s not a bug, it’s a feature

The Steelhawks and their management and coaching team did nothing wrong when they discovered the exploit. They were looking to give their team an edge just as pretty much every other GM is, and they happened to find a very effective way to do that that didn’t break any rules. You could even argue that it barely even counts as a bug or an exploit in the first place because it has been common knowledge for a long time that Scoring is generally more useful in the sim than passing. Everyone who has been here a while knows that, new members frequently get told to build their players in a shoot-first kind of way and 80-99 builds have never been seen as something that is particularly fishy until the standards changed recently.

All the Steelhawks did was take this a little bit further than what everyone else was already doing. Looking back now, it’s kinda surprising that it even took that long for someone to discover this. You could even go so far as to blame the rest of the league for not being smarter and drawing the obvious conclusion of “Lower passing than scoring seems to work better, so why don’t we try lowering passing even further?” way sooner.

2. Not all test-sims

“Test sim” seems to be the term of the hour, because basically all these discussions we’ve been having about the exploit have largely been based on various test sims. Test sims that have been done by the Hamilton management and by independent members, test sims from the STHS forum with NHL-builds, test sims with actual SHL builds and experimentally manipulated ones and so on. There is a plethora of test sims available at this point, the problem is that they’ve been coming to some vastly different conclusions and we still don’t have a place where all those test sims are collected for us to compare the results.

A large portion of the criticism aimed towards the Steelhawks is based on just a handful of tests that were done back when the exploit first became public knowledge. There were a few tests and Screenshots out there that showed players with the 40-99 build absolutely destroying the league and the individual scoring lists, which basically convinced the public of how bad this whole thing is. The problem is that since then, many other tests have come out that had very different results and came to very different conclusions. Tests that showed a slightly higher win-rate for those high-gap teams but not a super significant one, tests that simulated a high-gap players career numbers and found that they weren’t really higher than those of regularly built top-players today. But those tests, which both would’ve calmed the discussion and absolved the Steelhawks somewhat, didn’t nearly receive the same attention as the flashy ones with the crazy numbers.

I’ve said it before and I’ll say it again. The data base that this whole discussion is built on is much more shaky and inconclusive than people are willing to admit. The tests that show a big effect are prominently featured, the ones that don’t are largely ignored. And then there also are a bunch of other tests that no one outside of the office has seen yet. Making all those public would go a long way in accurately assessing this situation.

I am not saying that I don’t think this exploit exists, what I’m saying is that, contrary to what many people claim, we don’t have a lot of proof for it yet, at least not proof that shows more than a relatively minor effect. Maybe that proof exists already in the various test sims that the HO has conducted but if so, we haven’t seen it. Instead we chose to point fingers at Hamilton based on a couple of tests, but ignore a lot of the data that would absolve them.

3. 50 > 15

Now all the talk about data and test sims is good and well, but how about the actual results of this season? After all, Hamilton seems to be well on its way to win the Cup in sweeping fashion, does that not prove how well their exploit is working? Well, no it doesn’t. A dozen or so playoff games don’t prove anything from a statistical standpoint. There are so many factors that could be at work here. We all know how random STHS can be, teams go on Cinderella runs or unexpectedly falter in the playoffs all the time and in such a small sample of games, basically everything can happen. Even if Hamilton ends up sweeping the finals, that doesn’t prove anything. Be prepared for me to talk about sample size again and again over the course of this article, as it is basically my main argument (and the only one I really need) about people who are condemning Hamilton based on how the playoffs are going.

The only bigger sample we have at 50 games is the regular season, which is still subject to a lot of randomness, but is a better data base than the 15 or so games they had in the playoffs. And during that regular season, Hamilton did well but they didn’t break the league. Hell they didn’t even finish first in the standings but finished nine points behind the first-placed team and two points behind the team they are now facing in the finals. Is this exploit some weird kind of magic that only works in the playoffs but not the regular season? I’d say that’s highly doubtful.

4. Goalies, Scorers and the Playoffs

Let’s get into the statistical weeds some more by looking at a couple of things: The playoffs, goalies, the league’s top-scorers and Hamilton’s roster. Let’s start with the playoffs which, as I already said, shouldn’t be overinterpreted anyway, but who still offer some interesting observations. So if you look at Hamilton’s road to the finals, who did they play against? The first opponent was Buffalo, a wildcard team with an 800 TPE goalie. It’s really not all that surprising to see that series end in a convincing Hamilton win on these two facts alone. Now if you look at Buffalos roster you will see that they are actually a pretty good team, better than Hamilton and a bunch of others in these playoffs even, at least on paper. On paper though, that’s the key thing. Buffalo has struggled all season to live up to their roster’s potential, why is it so surprising to see these struggles continue in the playoffs?

After that it was on to the Rage, a series which Hamilton would win 4-1 but after a bunch of pretty close games. Manhattan would end up losing a game were they heavily outshot Hamilton, then winning one where they were outshot themselves. Typical STHS playoff-randomness I would argue and certainly not an outrageous result. Oh, did I mention that Manhattan switched goalies mid-series because they didn’t have a clear cut #1, but two mid-range TPE goaltenders, one at around 1100 TPE and one at around 900? That probably didn’t help them either and brings me to the last series, which is still ongoing.

Because here, in the series against Edmonton, we once again see a common theme. In the first two games both teams outshot each other once, but Hamilton won each of these games maybe because they have a 1700 TPE goaltender going up against a 1150 TPE one? Here we have a team with one of the highest TPE-goalies in the league who have yet to face a goalie with 1200 TPE or more in theplayoffs. With this big of a TPE-difference at the most important position in the game, is it really that surprising that they are winning a lot of close games? Now, I will acknowledge that there is still some weirdness there. Edmonton’s lower TPE goalie was fantastic in the regular season and Hamilton higher TPE one wasn’t very good, whereas now in the playoffs, Edmontons goalie saw his performance level slightly drop whereas Hamiltons goalie saw his save-percentage jump from 0.904% to 0.934%, which is a crazy improvement. Is this just randomness again or some kind of effect from the exploit? I don’t know but once again, I will argue that it would be pretty weird if that exploit activated during the playoffs but not during the regular season. Maybe it actually does but if so, we are still well on our way to gathering proof for that and it would mean that there are other dimensions to this exploit that haven’t been taken into account at all so far. Which brings me back to my previous point: We need more and better organized data to be able to legitimately draw these conclusions.

One quick last word about Edmonton by the way. Maybe some of you have already noticed based on this CW’s trivia question, but the Blizzard made it to the finals on a roster that is over 50% pass-first players! So apparently, having a team that basically does the exact opposite of what this exploit demands still works, it can certainly make you the second best team in the regular season and carry you into the finals even without a super high-TPE goalie. It almost seems like in these finals, we have a clash of the systems, a clash of philosophies: Shoot-first vs. pass-first. Looks like right now shoot-first is winning but over the course of 3 or 4 games, everything can happen. If you look at the regular season, the exact opposite was true by the way. Edmonton, the pass-first team, won both games against the shoot-first Steelhawks with a score of 4-1. So once again we have a small sample size that isn’t really fit to draw conclusions on and even if Hamilton sweeps the finals on the back of a dominant goalie-performance, the season series between the two teams would still be just 4-2, with the Steelhawks merely winning their games when it mattered more.

Now all this talk about goalies and their performances brings up another point of course. Maybe having a high-scoring, low-passing build simply makes it so that your players shoot better, making every opponents goalie look worse against you than he would against other teams. I can’t say too much about that claim other than that I have never heard or seen any evidence of it being true. Ever since I came here, people have been saying “Higher scoring doesn’t make your player shoot better (or only slightly so), it just makes him shoot more.” Who knows, maybe that decade old piece of STHS-dogma is wrong? It could very well be but once again, it would mean that we are still in the very early ages of collecting proof for it and it would once again add further substance to the notion that we don’t nearly understand this exploit well enough and that much more evaluation and data is needed instead of finger pointing.

Another possible statistic to look at are the individual scoring lists of the recent past, after all it were crazy high individual scoring numbers that got people sold on the severity of the exploit after the first test sims. If you look at those stats, you will find that for each of the last three seasons, the leagues overall top-scorer was a pass-first player. For the past two years it was Mikhail Lokitonov and this season it was Tony Pepperoni and both haven’t exactly been flukes but players who have consistently done well for the past few seasons. In Season 51, the scoring top-5 consisted of two pass-first players, two regular shoot-first players with a gap of around 10, and only one high-gap player (Phelps) who a) has been one of the best forwards in the game for a long time, and b) is actually one of the higher passing players on a team were a lot of players have similar or higher gaps. And if you look at the S51 playoffs (which isn’t super useful, I know, because of the different numbers of games played), you will find that the top-5 consist of a whooping four pass-first and only one shoot-first player.

Also, when you look at Hamiltons lowest passing players, some did really well (Wilson, Jazz), some did okay (Johnson, Miller) and others did very little (Fouquette, Owens, Michaud, Anazibf). Unsurprisingly the high-TPE guys usually did well and the low TPE-guys didn’t, with Jazz being the only real exception who outperformed his TPE-range. That’s nothing unusual, that’s how things will look on pretty much all teams. And if you look at the teams top-performers, you will find more players with a gap in the 15-20 range than those with 25-30. Of the teams five players with a 25+ gap, only one did really well and finished second in team scoring: Aaron Wilson, the guy who is their 2nd highest TPE-forward anyway.

5. It's a rules thing

Like the Head Office and their approach or not, you do have to acknowledge that they acted quickly when news about this first came out. They definitely need to do more to robustly solve this problem for the future, but so far the band-aid they put on the problem last off-season has largely worked. Reverting the redistributions and build changes that Hamilton tried to do already severely limited the possible effects of the tactics than making their builds much less extreme than they could have been. The Head Office should have implemented a maximum PA-SC gap of let’s say 15 points and prevented the regression of inactive players to make them high SC, low PA. But the fact that they didn’t do that is their own fault though and you shouldn’t blame the Steelhawks for the shortcomings of the Head Office when trying to deal with this problem. They are operating within the rules and it is our fault for not tightening the rules further when we had the chance.

Alright, see you back in a bit when I will be making my case AGAINST the Steelhawks...

Words: 2500

***hotdog*** · 01-10-2020, 12:01 PM

01-10-2020, 11:50 AMRomanesEuntDomus Wrote: 3. 50 > 15

Now all the talk about data and test sims is good and well, but how about the actual results of this season? After all, Hamilton seems to be well on its way to win the Cup in sweeping fashion, does that not prove how well their exploit is working? Well, no it doesn’t. A dozen or so playoff games don’t prove anything from a statistical standpoint. There are so many factors that could be at work here. We all know how random STHS can be, teams go on Cinderella runs or unexpectedly falter in the playoffs all the time and in such a small sample of games, basically everything can happen. Even if Hamilton ends up sweeping the finals, that doesn’t prove anything. Be prepared for me to talk about sample size again and again over the course of this article, as it is basically my main argument (and the only one I really need) about people who are condemning Hamilton based on how the playoffs are going.

The only bigger sample we have at 50 games is the regular season, which is still subject to a lot of randomness, but is a better data base than the 15 or so games they had in the playoffs. And during that regular season, Hamilton did well but they didn’t break the league. Hell they didn’t even finish first in the standings but finished nine points behind the first-placed team and two points behind the team they are now facing in the finals. Is this exploit some weird kind of magic that only works in the playoffs but not the regular season? I’d say that’s highly doubtful.

This is absolutely true, and I'm glad people are talking about sample size - please keep stressing this, I'll be right there with you. Claiming that 15 games provides proof of a broken system is asinine. Unfortunately for Steelhawks supporters, tests on seasons of a much larger sample size (600 games) prove the same point. I ran loooottttsssss of those 600-game season tests this season and Hamilton cruised to the top of the East every single time (literally every one). 15 games isn't proof that the tactic works, and 50 games isn't proof that the tactic works, but 40x600 is stronger evidence. Feel free to use that as evidence in your "against" column if you wind up writing it Smile

Great stuff RED!

TheWoZy · 01-10-2020, 12:06 PM

smh

***artermis*** · 01-10-2020, 12:18 PM

Only thing I disagree with is the notion that if a GM does whatever it takes within the rules given to better their team then nothing wrong can be done, whilst I believe that morals and ethics are always present and hold higher standards than the rules, or are the only guideline in a situation without rules.

Now, I'm not saying what Hamilton did is still inherently wrong, I'd say it matters more so on the interpretation of the team at the time: was it more so that having a large gap helps the players perform better, just as having high defense does? Or was it that STHS had a hole that could be exploited? The viewpoint of those spearheading the effort in Hamilton is key.

**RomanesEuntDomus** · 01-10-2020, 01:47 PM

Part Two is out now!

Tomasnz · 01-10-2020, 06:11 PM

good shout on the goalies, only thing to note there is that goalie builds max out at about 1250TPE so there is some advantage in 1700 over 1100 but not as much as for skaters (who max out 1500ish)

Navigation

Extra Menu

About us