Who will benefit in 2020

Who will be at an advantage and who will be at a disadvantage with the changes for 2020

Superpesis announced on Friday that the format for the 2020 season, which has been severely hampered by the Coronavirus. In a change from last year, there are 4 key differences:

1. No relegation - this was already public knowledge right from the outset;

2. The season is shortened to 24 games;

3. The top 3 teams get to choose who they play in the playoffs (which are shortened series); and 

4. The format of the fixtures is such that some teams will not play each other in the regular season. 

The last of the 3 changes is really dramatic and has been arranged along relative geographical lines. In short the teams are split into groups (North, South, East and West) with the North and South groups containing 4 teams each and East and West groups with 3 teams each. Each team will play ever other team in their group 4 times (2 home and 2 away) as well as the 2 “adjacent” groups a further 4 times. In layman’s terms, each opposite group will not play each other, i.e. North group plays everyone except the teams in the South group, East plays everyone except West.

Once you get used to this idea, you start to wonder how the changes will impact the outcome of the league, for example if Vimpeli and Joensuu aren’t playing against each other in the regular season, will they take home more points in games against lower table opposition?

I have spent the rest of the day thinking about this question and undertook 3 different mathematical and theoretical models to explore who might benefit and who might suffer from these changes.

The Exclusion method

So what would have happened in 2019 if we disregarded results between teams that will not play each other in 2020 under the new rules? By excluding last years results, taking an average points per game and multiplying by 24, you get a sense of what the table could have looked like. This method is rather crude and has no real mathematical simulation to it. But in a way that is why it is quite useful; by excluding real results from opponents who a team wont play in 2020 and giving greater weight to results from games a team will play, you can understand the power dynamic between the teams a little better.

So what were the results? Well, the top 4 would still have been pretty much the same, but there are a few changes lower down which were worth noting. Hyvinkään Tahko would have fallen to 7th on this model, with Seinäjoen JymyJussit and Pattijoen Urheilijat making 6th and 7th respectively. Equally the playoff picture would have been different for Kiteen Pallo, who would have lost out to IPV under this model. The final observation is that Siilinjärven Pesis would have been in 11th place and Kempeleen Kiri would have bee in the initial relegation playoff with 14th place. 

If this model shows a trend for where some of the strengths or weaknesses of teams were against opponents from the other side of the country, then I would expect SiiPe and IPV to be net winners from the changes, with a noticeable loss for Tahko.

The adjusted Pythagorean method

Throughout 2019 I monitored teams performance against a modified Pythagorean prediction formula that works on runs for and against being indicative of how often a team is likely to win. For this calculation, I modified the expected number of runs for and against each team based on who theory opponents will be in 2020. In this method i included newly promoted Manse PP to ensure the formulas balanced. I wont go into detail on my workings (and I am certain I could fiddle with the formulas for many more hours) but it is far too technical to cover in a short post like this one. The method looked at the differences between the expected win:loss ratio when the 2020 format was applied and what it would be if it was not applied.

The findings for this one were more nuanced, but suggested that the top 4 teams from 2019 (Vimpeli, Sotkamo, KPL and Joensuu) would benefit. This makes sense when you think that they will not have to face a key rival. Overall however the groups seem fairly balanced, with teams generally performing as they did in 2019. Once again however, Tahko seemed to slide from 5th and give the spot up to Sienäjoki. But in this model Kitee prevail over IPV in making 8th place in the table.

Modified Power ranking model

This model works on the basis of taking each teams modified points total from 2019 and 2020’s Halli SM games and modelling them against the opponents they will play in the regular season. The power ranking model I use during the season basically looks at the importance a victory means (or even any points) against any particular opponent, based on how successful the opponent has been in that season - the more points an opponent has in the season, the more important the victory is. 

The results showed a huge advantage to the “top 4” teams mentioned above, but also showed a huge boost for PattU and Manse PP. The teams that seemed to lose out in this model were Koskenkorva, who came in at 13th and SiiPe in 14th place. 

Flaws in the modelling

There are some significant flaws in each of the models I have used because the work on a number of assumptions which are not strictly true - for example the team in 2019 would play the same way in 2020, there is no margin for new signings, for players ageing or becoming better players. One other major impact of the changes for 2020 is the lack of relegation - this could fundamentally shift the attitudes of teams as to how they approach games when they are out of the running for a playoff spot, but know they are safe from dropping to Ykköspesis no matter what. But to be honest this was just a bit of fun to see what kind of impacts the change in the format for 2020 could bring.

What does my instinct tell me?

Ignoring the models above and just looking at the teams and their schedules, instinctively i think that there could be a gulf between the top 4 and the rest of the league. As lower teams rest up and protect players etc. The top 4 will be jostling for the right to pick their opponent in the first round of the playoffs. I also think that Manse PP will benefit with the changes, not having to play Sotkamo or PattU. Indeed, if teams at the lower end of the table do start resting players late in the season, Manse PP may look to take advantage and make a mark on the season.

I think that Kempele will be in trouble - the recent developments and changes with age manager only add to worries about 2020; but with nothing at stake if they have a difficult season, Kempele may be content to try and keep their head down low and look towards 2021.