Player relevance / usefulness

Hello everybody,

tonight we have released a new version of hattrick youthclub (hy). Besides several smaller bug fixes and improvements we introduce a new feature in hattrick youthclub which replaces the bad player detection:

On player details page and player skill matrix we show the relevance (or player usefulness) for every playerΒ to you. The player usefulness should help you (especially the beginners!) to figure out on which player it is worthy to pay a attention. Therefor we calculate the relevance a player has right now (dark bar) and the relevance he can get at max (light bar). We take a lot of aspects into account like current skills, max skills, top3 skills, allrounder skill, twins, rejected players, speciality, age, earliest day of promotion and current skill estimations.

We hope that you’ll like this new feature and we hope that every user, but especially the beginners, will benefit from it. For more detailes about the relevance please checkout our corresponding wiki page at http://wiki.hattrick-youthclub.org/en/hy/relevance.

Please note that the relevance rating may be empty for you now. To trigger a calculation just press the sync button on top of the page or just wait until your next match is played.

We are looking forward to your feedback,

Ruben (Mod-Homer-Simpson) and Markus (Mackshot)

16 Responses to “Player relevance / usefulness”


  • As you said except for being relevant to the beginners I fail to see the usefulness of this column. Can I have the option to remove it? I am the one in charge of my players and I actually dislike that a formula will tell me who is going to me my next best youngster. So for more experienced users allow us to hide this column please.

    • Hi Stephan,

      well also in the past some users complained about new features. Then we added an option to remove it, but in the long run less then 1% of the users checked that option. So a few month ago we removed all these options and surprisingly nobody (not even the person asking for the option to remove it) complained about it. I feel that this would happen again if we would add such an option. But maybe you are not the only person who like to hide it, so let’s wait for more feedback.

      But in general I dislike your argumentation, because 1) it is wrong regarding the meaning of the formula and 2) you still have to interpret the relevance together with the skills etc. So the formula does not tell you “do this and not that”.
      The formula does not tell you who is your “next best youngster” it just highlight the players you should focus on. There are different reasons for that, player skills are one, but another one is, that you simply do not know enough to decide if a player is good or not.

      Cheers,
      Markus

      • Hello,

        A “parameter” option for this bar will be great : if we could add/delete some aspects (ex. age, top skills, …), we could use this tool better.

        Currently, I dont’ know what it represents so I don’t trust it and I find it interesting, but useless.

        • Agree with your comments, but than this is something that I do in an excel file anyway and I do it only when I feel I have more than one-two very good players, which is not too often the case.

          For U17 scouts this tool could be interesting, to pick players fro a larger pool, use the age as a filter, etc., but for the likes of the average user, don’t think so.

        • Did you checkout the wiki page? There we name all factors and describe very briefly how we use the factors. May this helps.

          • I checked it but it doesn’t clear for an external user. If I haven’t control on the parameters, I cannot trsut him, some players seem to be overestimated ; so I guess some others can be underestimated.

            For instance, this player : http://www.hattrick-youthclub.org/site/player_details/player_id/3744122

            17.89yo, no great skill and yet he has a 55% potential.

          • Well as the wiki states (please read it!) we do take into account with which age the player could be promoted. What is not explicitly stated there is that we do not take into account the current age and the training progress. This is impossible due to several reasons.
            Please also read my reply on the bottom of the page, there I will state 2 other aspects, which could be interesting for you, because you are one of the user who brought that up, thanks!

  • First off I love this idea and I wholeheartedly endorse it.

    However the goalie relevance is a little confusing. It seems like a lot of players who clearly are not goaltenders (and who I’ve even fielded as a goaltender) are still being shown as ‘relevant’ with the blue bars. Maybe I’m not understanding it correctly, but how do you get rid of the blue bar if a player is clearly not relevant to the position?

    • Thank you for your feedback. Can you please name a conrete example of a player where you feel that there should be no blue bar at all and why? Then I can have a close look and may can explain it you or fix it, if it is a bug.
      Thank you!

      • youthplayerid = 140411744

        He has a 0%-82.2% relevance for keeper for me, but I think he’s clearly a (bad) fielder.

        • Maybe you named a bad example, but I do not see why this player can be no keeper. You have no skill information and no star information about the keeper skill / position. Please explain it to me or if I am right and you picked a bad on, please look for a better example πŸ˜‰ .
          There is no indication that the player’s third top3-skill is not goalkeeping with 7 and defending could also be 4 as well. Then it would be a good goalkeeper πŸ˜‰ . The chance is low, but the chance that a player is a goalkeeper is always pretty low.

          Thanks,
          Markus

  • I equally do not find this so useful. I would rather prefer to have a column where to display other more interesting info like: a player being reported as the one with most potential, other interesting info from the reports or from games, like how many times a player was selected as a captain or as set pieces taker, etc.

  • First, please let me underline that we have never said that this relevance rating is perfect. It can help beginners, but can also help advanced users to ask themselves if they may focus on the wrong player or not. It is like a helping hand and not a god-mode πŸ˜‰ .

    We mentioned in the wiki that we are aware of the fact that multi-skillers are overestimated. Yesterday evening we got an idea how we can lower this overestimation (but we will most probably not get completly rid of it). We will implement it and will see if it does fit better.
    The formula for the relevance will never be perfect and is “work in progress”.
    We really understand that some people are not sure if they can trust in it. Thus I got an idea last night: We might publish the sourcecode for this formula on github. Thus everybody can have a look. If somebody does not understand the formula he can ask somebody who have some more experience with programming. You can send in suggestions for improvements via github and we can discuss about them. We feel that this help the users to understand what we do, to understand why it is not easy and users can to help us to make the formula better. What are your thoughts about that?

    • 1. The idea is ok, but the raw formula isn’t intuitive for “normal people” : I’m pretty sure that a lot of them never look at it and simply ignore the bar. A tool that allows players to introduce some data will be better (like TSI Keeper or HO Scout for instance).

      2. A great thing will be the “player edition”. Currently, we can change some datas about our players, but I found them pretty useless. I’d like an advanced editing possibility that allows me to add some infos (current level / maximum level for my players) with a different colour. The twins system does it, so i guess that it’s possible. The star calculation is pretty easy to understand, so the users could add some datas to their player with this option.

      • Hi mazko,

        about 1.: Sure it isn’t, but people who are (pretty) motivated (and skilled enough^^) can help us to improve it and can also explain some basics to others. I think it is a win-win situation. And thus everybody can decide if he believes into this formula or not.

        about 2.: First of all: This is a complete different topic. I am not sure if we should implement it, because I do not know whether many people would use it or not and how it should work exactly. But we are very open to any idea, so I really like to encourage you to take some time to think about your idea and all consequences. If you come to the conclusion that it is a good one, please describe the idea in detail (how it should work, what should be calculated, what is the user input, etc) and send your idea to our email address. We will come back to you with our questions and thoughts about it. And may we like it and implement it and send a big thank you to you πŸ™‚ .

  • We have updated the calculation tonight. We have lowered the relevance for Multi-Skillers and for players who only have a high passing potential.
    As mentioned on facebook and twitter, we will work on it time by time and may will also make our formula public, so that you can help us.
    We are aware of the fact that the relevance formula will never be perfect πŸ˜‰ .
    I am leaving for a short vacation tomorrow. So all next steps will not be made before I will be back πŸ™‚ .

Comments are currently closed.