Recursive Recruitment
Uncommon · Sorcery
6.0PlayableRank 108 of 280
61.0%Rank 24 of 271 by win rate, card in hand · set average: 56.2%
How are these measures computed?
Pick rating. The pick rating estimates what a card is worth when you choose it, before there are results: it combines several signals (expert evaluations, the choices of strong players, characteristics of the card such as its cost and rarity) and was tuned on the results of decks of strong players. It is for comparing cards while you draft.
Observed strength. The observed strength is the win rate of the games in which the card was in hand at some point (opening hand or drawn), from hundreds of thousands of real games of online draft players since the set came out. The rank compares the cards seen in at least 1,000 games. It does not correct for the strength of the deck the card is played in, and a result stays uncertain for a card that is played little (about 1.6 points of margin at 1,000 games).
It wins more than its rating suggests.
Why this gap?
The pick rating is a prior: it is for choosing a card before there are results. The observed strength comes from real games, and the win rate of a card is stable from one week to the next. The gap often comes from a card that is easy to replace at the pick but very efficient once played, from a good archetype that carries it, or from a flaw of our rating: in a set that already has many games, the results are more reliable than the rating.
Good playable that always makes the cut (6.0/10).
Why. Two 2/2 Cadet tokens for four mana, plus a late flashback that makes them bigger depending on your graveyard. A good source of board presence and late-game value.
Risks. It costs four mana across two colours. The flashback costs eight, so it only matters in long games, and the tokens are fragile against sweepers.
In the draft. Good in a blue-black deck, which is its main pair. Take it fairly early if you are in those colours; it matters less outside them.
Win rate of the card
Percentage of games won by the decks that play this card, depending on when it shows up, from the real games of the set.
Gap when it is drawn, against the games where it is not : +6.3 win-rate points
Its place in the set
- Uncommon: rank 39 of 109
- Multicolour: rank 34 of 54
- Cost 4: rank 16 of 37
- Average pick: 4.0 (1 = taken first from a pack; the lower the number, the earlier players choose it)
Win rate by colour pair
Percentage of games won by the decks of each pair that play the card, with the number of games. The pairs are the two main colors of the deck: when the card has a color outside the pair, the deck plays it as a splash (marked “splash”). A pair that has an archetype in this set leads to its guide.
splash
splash
splash
splash
Win rate: strong players and all players
Percentage of games won when the card is in hand; strong players are those ranked diamond or mythic online.
Win rate on the play and on the draw
Win rate by number of copies in the deck
- 1 copy57.5 % · 33,489 games
- 2 copies59.8 % · 9,111 games
- 3 copies60.3 % · 1,479 games
- 4 copies or more62.1 % · 140 games
Win rate alongside other cards
Win rate of the decks that play both cards, compared with the one you would expect from what each card gets alone. At least 150 games together.
Cards it wins more with
- Frostbite Pyromental198 games · 58.6 % (instead of 54.9% expected)
- The Echoverse Fulcrum1,295 games · 62.8 % (instead of 59.8% expected)
- Terminal Criticism491 games · 56.6 % (instead of 54.0% expected)
- Loot, the Anomaly931 games · 55.5 % (instead of 53.2% expected)
Cards it wins less with
- Darklight Phoenix709 games · 50.1 % (instead of 56.2% expected)
- Guiding Hydra925 games · 54.2 % (instead of 58.4% expected)
- Rampart Hunter858 games · 48.0 % (instead of 52.2% expected)
Other cards of the same cost and colours
2. The Theorist, Jace Beleren
9.3
3. Denzilore Fatehold
9.1
32. Sphinx of False Conclusions
7.4
37. Darklight Phoenix
7.2
Intrinsic ratings: they do not take into account the cards you have already picked. They matter most for the first pick of the first pack.