An unpredictable story of an unpredictable player

It was only a month ago when a user asked in the “Questions to HTs” thread in the Global forum about the strange behaviour of Unpredictable players during the Own Goal special event. The users’ own research showed that the frequency of Unpredictable Own Goal was inversely proportional to the amount of unpredictable players a team has.

Almost right away, a discussion started between users if this was by design or actually a bug in the code. The main problem was that the higher the number of unpredictable players in your own team, the higher the chance was for your opponent to receive an own goal. Of course, this was not our intention when we designed the new special events and the reason for all this was entirely down to a bug in the code. We released a fix right away, but as a result, the specialty of Unpredictable players was nerfed significantly, since all those negative events now hit your own team instead.

After that, a long discussion started in another thread in Global , trying to find a way to minimize the negative effect without a significant impact to the rest of the specialties.

In the meantime, during the tests, a similar bug was identified in another negative event, the Unpredictable Mistake. This time, the correct team was assigned the event, but when the match engine checked if the event would actually be triggered or cancelled, it erroneously would check the amount of players used in the opponent team instead of your own.

As we explained in a previous blog post, the chance to create a new special event during a match is interconnected, depending on all the available specialties in the field. Lowering the frequency of one event will immediately means that we increase it for another one. Thus, we had to be careful and avoid any big change that will alter the plans users have laid down based on the de facto reality that have seen in the game.

This coming Monday, June 22, a fix will be released in the match engine that also tweaks the other specialities. We will present here the difference between the two versions. To do that, we introduce the concept of the theoretical “expected goals” value. This value shows how many goals you should expect in a match for each special event. However, there are several other attributes that affects the actual result – such as the players’ skills or the specific lineup of your team and your opponent’s team.

The expected goals used in this blog post are meant to demonstrate the impact this change will have, and should not used to base your decisions on all by itself.

In the following table, we see the expected goals for the affected special events, before and after the changes in the match engine. The values were calculated based on data from real matches from a variety of different lineups and skills sets.

As we see, the only affected specialties are the Head and the Unpredictable in the above four special events. In the first column, we see the values of the four events before we do any kind of calibration. The second column is the one that will be released next Monday where we do the following changes.

• Move some of the Own Goal frequency to Corner to Header
• Move some of the Unpredictable Mistake frequency to Winger to Header
• Reduce the success ratio of Own Goal
• Reduce the chance to have more than 1 of the above negative events in a single match.

In the following table we see the success ratio before and after the changes for the Unpredictable Own Goal event, where the passing skill of the Winger/Forward and the goalkeeping of your Keeper is important. It’s clear that it’s significantly harder now for a player to score an own goal

This short project was not a reason for us to touch the rest of the special events, which could have created a disorder in users plans without further announcement. We may do a more generic re-calibration in the future. No such revision is planned today, but if we do one we will as always inform users in advance.