04/08/2014

The science of underevals

"Underevals", or players that gain skill bar on transfer, are starting to become a more popular method of farming players, particularly with the new threads in the forums created for the purpose of advertising players. In a previous post, I discussed what happens when a transfer takes place, but I'll build on that by discussing how you can use that to your advantage when selling players.

The basic premise

Players gain or lose skill on transfer, based on their average form, match experience and the amount the season has progressed. The way the skill increase / decrease is calculated is very similar to the way that it is at the change report, although there is one difference; at the change report, players get a random amount of between 0 and 0.2 as an additional gain (or loss) at the CR, but this is not applied on transfers.

Setting players up to become underevals

Building players up in order to sell them as underevals is a similar approach to training players normally - maintain form and ensure you give enough game time. The higher the form and the greater the match experience, the bigger the skill increase. Selling later in the season will also increase the skill gain a player makes. It's worth noting, however, that you don't necessarily need to get a huge skill increase - just enough so that the player gains a skill bar. You may find that, if you have a player that is high in their skill band, it's easier to try and sell them as an undereval than trying to attain a double jump with them, especially if you have lots of other players that need match experience.

When exactly to sell

There are a couple of things to consider when trying to sell underevals - whether you've earned enough of a skill increase for a visible jump on transfer, and whether there are teams actively looking for players. While selling four days before the deadline will give the maximum increase, you won't want to sell then if there aren't many leagues in silly season (which you can spot in the official leagues, but is mainly down to luck in private ones), and there's not much point keeping a player for extra weeks if you've already done enough for them to gain skill.

Calculating the DV a player needs to be an undereval

This calculator should help you work out what DV you will need to make a player an undereval given a point in the season. Note that I have written the calculator to be slightly pessimistic, in that they should return a DV that will guarantee an undereval, but you use it at your own risk.

of days / games

7 comments:

  1. Confused a little with numbers i'm getting for a player.

    Age 16-20
    skill 7.7
    average form 16
    6/10 games

    19DV gives 0.9 skill

    same player but average form 18

    20DV gives 0.9 Skill

    does that make sense?

    ReplyDelete
  2. It might not appear to, but the calculation should hold.

    Because of the point in time you're selling, to get 0.9 skill is going to require you to have a season end CR of 0.9 * (10/6) or 1.5. The 0.2 random element does not happen with a mini-CR.

    The only DV value that will guarantee that (according to my recorded CR data) is 20. For the player's DV to extrapolate out to 20 at the end of the season, the player would need an astronomical amount of match experience - in this case, enough to yield 75% "bonus" (i.e. 75% of 20 - their AF) after 60% of the season.

    For a player on 18.0 AF, this would mean they'd need 19.5 DV, which rounds to 20.
    For a player on 16.0 AF, this would mean they'd need 19.0 DV.
    For a player on 0 AF, this would mean they'd need 15 DV.

    I presume you're selling as late as possible in a 6 team league with 2 games a week. Attempting to get an undereval in that sort of format with that sort of required increase will be pretty much impossible.

    ReplyDelete
  3. yes that's right. I didn't think it'd be easy! To be honest I was really just playing around with the calculator post CR.
    Thanks for the reply. It still seems a tad odd to me but I get what your saying.
    Also thanks for the blog in general! I always give a link to anyone who is just starting the game and I honestly feel it's a big factor in keeping them sticking with xpert after they lose their first five matches!

    ReplyDelete
  4. Hi, I've put up 17/6 (real skill 5.8) who I was hoping to sell as an undereval. However according to your calculator I needed a DV of 17 to guarantee the 0.8, unfortunately I only managed a DV of 16 so I was wondering if there is any chance at all he's a 17/7 after evaluating? Or because of the +/- 0.2 not happening does it mean he will definitely still be a 17/6?

    ReplyDelete
    Replies
    1. Hi,

      The DV that gives a guaranteed skill increase is used because DV is rounded i.e. 16 means anywhere from 15.5 to 16.5, and a "low" value may yield a smaller increase than a "high" one. I deliberately give the DV that guarantees an increase because I don't want managers turning round and complaining that I've told them they'll get a skill increase on transfer when it is only a possibility.

      It could be possible that the player has earned the bar increase, although I'd need to know the percentage of the season that has elapsed at the point of listing. Looking at the figures I have (which are available in my CR data post), I presume you're selling after 80% of the season?

      Delete
    2. It's in the Dream League so we're 12/14 games into the season. Although thinking about it I think I put him up for sale on Sat so will that mean only 11/14 will count? If so I've screwed up.... I've received a bid for him anyway so he's definitely off!

      Thanks for the reply.

      Delete
    3. Update, he did jump on transfer, guess I got very lucky!

      Delete