Here comes the first part of the $4 list. I look forward to hear your complains
I think this list will be the most discussed one, even more than the $5 list.
There seemed to be very difficult decisions, especially in the middle ranks you will see in the next part.
 | #43 Thief (Base) Weighted Average: 41.22 / Median: 42 / Mode: 43 / Standard Deviation: 2.4 Highest Rank(s): #34 (1x), #35 (1x), #38 (2x) / Lowest Rank(s): #43 (13x)
Thief is without doubt the worst $4 card in the game. The next card has a healthy margin and Thief has by far the lowest deviation of all $4 cards. Nearly the half of all players ranked it last.
An attack on the last position may seem strange, but Thief has the big problem helping your opponent in the early game. Its a free trasher for your opponent and even later it's so risky hitting the Coppers of your opponent. Its only use may be in thin Chapel decks or if you manage to play it multiple times per turn. And it gets better in 3- or 4-player games, where you can minimize the risk of getting nothing and hitting Coppers. It can be a nice counter against a Ill-Gotten-Gains rush dealing out curses with it, but it is still swingy and therefore still a bad card. You don't want it early and in the later game it's almost never worth a buy.
|
 | #42 Scout (Intrigue) Weighted Average: 39.12 / Median: 38.5 / Mode: 37 / Standard Deviation: 5.3 Highest Rank(s): #21 (1x), #28 (1x), #29 (1x) / Lowest Rank(s): #43 (4x)
Although it has a higher median than the next card and a pretty high mode (it got #37 6 times), it still only is second last. Only four cards have a lower deviation, so the consensus is still high and it even got last 4 times and #42 6 times too.
Scout has its uses. You don't need to spend an action, but it isn't a cantrip, so it really can hurt your deck. If you're massively greening this can be nice as it makes your next turn better, but is still not good. The best uses are: making Crossroads way better, it has a nice synergy with Wishing Well (making it a cheap Lab) and of course it's great with dual-type-victory cards like Harem, Great Hall and Nobles, making Scout a Lab or even better. There might be more edge cases when Scout shines, but those cases are rare and most of the times you can win without buying a Scout.
|
 | #41 Coppersmith (Intrigue) Weighted Average: 38.27 / Median: 39 / Mode: 41 / Standard Deviation: 6.2 Highest Rank(s): #16 (1x), #22 (1x), #26 (1x) / Lowest Rank(s): #41 (11x), #43 (2x)
Coppersmith is the third last card with a third of all players ranking it there. It has a below average deviation, but it's really high for such a low ranked card, because it has a few outliers in way higher rank regions.
Yes, Coppersmith is very hard to rank, because it's either clearly the worst card on the board or it's very dominating. As a opener you may get to $6 or even $8, but you also can draw only one Copper, so it's very swingy as a opener and gets worse later. On the other hand King's Court + Coppersmith can become brutal and it has some nice synergy with Apothecary and Counting House. The cases where it shines may occur more rarely than with any other card, but then it's a must-buy.
|
 | #40 Talisman (Prosperity) Weighted Average: 37.84 / Median: 38.5 / Mode: 40 / Standard Deviation: 4.2 Highest Rank(s): #28 (1x), #29 (1x), #30 (1x) / Lowest Rank(s): #42 (4x)
Talisman is the card with the third least deviation, so no surprise here. No-one ranked it last, but 4 people ranked it second last and 5 people ranked it exactly on #40.
There are very few cards for $4, you want in masses. Silk Road and Gardens may be a exception, but Talisman doesn't work with victory cards. So there are even less cards you want for free with Talisman. Caravan and in some cases Throne Room, Conspirator and Tournament came to my mind being the only cards which makes Talisman a good buy, especially as a opener. You can build a Village + Smithy/Envoy quicker too, but this is rarely worth a Talisman buy since you need money too. But it shines especially with cost reducers like Quarry, Bridge and especially Highway. Play Highway, play Talisman, buy Highway, get one for free, that's nice. Talisman is also nice for a quick 3-pile ending. This works well with Bishop for example: get many Talismans and then Bishops, trash Talismans for 3VP and try to three-pile. But Talisman also can hurt very badly since the free extra card is not optional, so only buy Talisman if you really want cheap card in masses. And don't forget: if you're buying more expensive cards you've spent $4 for a Copper.
|
 | #39 Bureaucrat (Base) Weighted Average: 35.75 / Median: 37 / Mode: 39 / Standard Deviation: 5.3 Highest Rank(s): #22 (1x), #26 (2x) / Lowest Rank(s): #42 (4x), #43 (1x)
We make the next big jump and can observe a close battle for not getting into the Bottom 5. Bureaucrat loses, maybe it got bad rankings more often with one last place, so that its high ranks couldn't save it.
The attack of Bureaucrat is weak. Your opponent loses one card that he don't need anyway for him getting another 4 card hand in the next turn. And he might even be able to counter that easily by playing Farming Village of Cartographer for example. The attack gets better in multiplayer games, especially if there are dual-type victory cards like Nobles or Harem on the board. The benefit on the other side still isn't good either. Top-decked silvers are nice, especially in the beginning and you can get to $8 with 4 silvers too, but it's not easy. So it seems Bureaucrat is nice where you don't want to get to $8, like in Duke / Gardens / Silk Road games, but then Bureacrat is a good counter too. And you can play your Bureaucrat less frequently if your deck is already flooded with silvers. Bureacrat + Big Money is not bad on the other side as it doesn't seem to have synergies well with other cards.
|
 | #38 Noble Brigand (Hinterlands) Weighted Average: 35.71 / Median: 36.5 / Mode: 41 / Standard Deviation: 5.6 Highest Rank(s): #24 (1x), #25 (1x), #26 (2x) / Lowest Rank(s): #41 (5x), #42 (2x)
Noble Brigand wins against Bureaucrat with a 0.04 point lead. It didn't got any last place, but got third last 5 times in exchange. So the order of these two isn't fixed.
Noble Brigand is the better Thief on many boards. It hasn't the disadvantage of trashing the opponents Coppers, making it a better opener. It even attacks on-buy. It deals out Coppers too which is nice playing against no-treasure decks. And it gives $1 too, so you have at least an immediate benefit. But it's worse in Colony games as it cannot steal Platinum, it cannot steal Ill-Gotten-Gains like Thief and cannot steal other Kingdom treasure cards. And most important: it's still too slow and doesn't hurt enough if you can't play one nearly each turn.
|
 | #37 Navigator (Seaside) Weighted Average: 34.56 / Median: 35 / Mode: 32 / Standard Deviation: 5.7 Highest Rank(s): #21 (1x), #22 (1x), #29 (1x) / Lowest Rank(s): #43 (4x)
Navigator is the card with the second most last places shared with Scout. But Navigator is the first card with a way higher mode with 7 times #32. It also has a worse unweighted average than Noble Brigand, so better players seem to rank Navigator higher than Newbies.
Scout is at least non-terminal and therefore nice for Wishing Well to draw the top-decked cards. Top-decking the next 5 cards in a specific order is only nice if you have still an action left to draw a few of them. Because if you don't do that, you draw all 5 cards no matter in what order you put them back. The discarding option is nice to minimize shuffle luck and to get a half Chancellor effect, but still it is terminal and most of the time there are better terminal cards on the board. At least it gives you $2. The best use still may be to enable Tunnel's reaction.
|
 | #36 Nomad Camp (Hinterlands) Weighted Average: 33.92 / Median: 33.5 / Mode: 32 / Standard Deviation: 5.3 Highest Rank(s): #21 (1x), #25 (1x), #27 (1x) / Lowest Rank(s): #41 (2x), #43 (2x)
Nomad Camp has two last places too with 4 last places still to come. It has a pretty low deviation, the fifth least like Scout. With only 4 times #32, the mode has not much significance here.
Woodcutter was the third worst $3 card. Here we have a Woodcutter with a on-buy top-deck ability. Is it worth costing $1 more? And how can it be that it ranks higher than Woodcutter? IMO there are only 2-3 reasons for that. You have a high chance to get a $5 card on turn 2 even with a 4/3 opening which is nice especially for Hunting Party and Cursers. Similar to that it is nice you need the +Buy either way and need multiple cheap cards as fast you can. But the only opening which is strong is IMO Nomad Camp / Fool's Gold / Fool's Gold. The last reason is if you're really unlucky in the late game and only get $4 and want to maximize the chances to hit $8 in the next turn. Beside of that it's only an expensive Woodcutter.
|
 | #35 Walled Village (Promo) Weighted Average: 32.86 / Median: 32.5 / Mode: 30 / Standard Deviation: 4.7 Highest Rank(s): #27 (3x) / Lowest Rank(s): #40 (2x), #42 (2x)
Walled Village is the next promo in the list. Its low deviation (fourth least) shows the consensus in its rank. As we get nearer to the middle ranks, the mode gets less significance, it was #30 4 times.
It is the worst of the four $4 villages. Why? Its only ability is top deck it if you weren't able to use both actions. So this is nice if you have only 2-3 terminal actions and really want to play them each time without taking the risk of colliding. The only reason opening with Walled Village could be if there's a $3 key card and you want a second one later too. Because of that Walled Village / Masquerade and Walled Village / Ambassador are Level 4 openings rankings on #91 and #114 on Councilroom respectively. In all other cases it is the same as the normal village.
|
 | #34 Pirate Ship (Seaside) Weighted Average: 32.71 / Median: 33 / Mode: 33 / Standard Deviation: 8.6 Highest Rank(s): #6 (1x), #9 (1x), #19 (1x) / Lowest Rank(s): #40 (4x), #41 (1x), #43 (1x)
Pirate Ship is only 0.15 points higher than Walled Village and has a clearly higher deviation than all cards before, but is still only 14th in the deviation ranking (that shows the big discussion I expect in the next part). It got last once (with three more to come) and is the first card that even got in the Top 10 twice.
There's the third attack that trashes treasures (tongue twister). Pirate Ship is the best one, but still not good. Or is it? The high ranks may result from players mainly playing 3- or 4-player games where Pirate Ships can be devastating. In 2-player games it's too slow most of the times. The advantage over Thief and Noble Brigand is getting virtual money and therefore not clogging your deck so you can play Pirate Ship more often and get the coins more often. So, with Pirate Ship you really want to buy as many as you can, so you can play them multiple times, and with Throne Room or King's Court this card is really great.
|
 | #33 Spy (Base) Weighted Average: 32.17 / Median: 32 / Mode: 31 / Standard Deviation: 6.7 Highest Rank(s): #17 (1x), #18 (1x), #23 (1x) / Lowest Rank(s): #42 (1x), #43 (2x)
Spy got on the last place twice, so there's only one more last place left. It has two outliers in the Top 20. It's unweighted average is lower than Pirate Ship, so higher ranked players seem to rank Spy higher than Pirate Ship.
An attack that is a cantrip, that seems nice at the first look. But Spy is an attack with a pretty bad attack and little benefit. It's very swingy as you can discard your victory card (or even your Tunnel) and discard the only Witch of your opponent, but you can hit a victory card of your opponent too that you put back. That's no change for your opponent and he even may use that additional info for the next turn. You can add a Spy in your drawing engine if you have a buy and money left, but is really rarely worth a buy. It may have ranked that high, because it's a cantrip and at least don't hurt (or hurts less than a few other cards already mentioned above may hurt). If we already cards with different costs (Warehouse and Cellar), the way better Cartographer comes into my mind.
|
 | #32 Treasure Map (Seaside) Weighted Average: 31.57 / Median: 31.5 / Mode: 31 / Standard Deviation: 8.8 Highest Rank(s): #7 (1x), #9 (1x), #14 (2x) / Lowest Rank(s): #39 (3x), #40 (1x), #42 (1x)
Treasure Map's stats seem similar to the Pirate Ship ones. Two outliers in the Top 10 too, similar high deviation, but no last place this time, only a second last place this time.
I can already here the outcry of a card that ranks that high and is so luck-dependant. Its power is undeniable thou. An early enabling can already decide a game. But you can hardly call it strategy going for Treasure Map. You really need enablers for that, like Warehouse, Chapel, Tactician or the Watchtower/Talisman combo. If you go for Treasure Map without such enablers, you totally rely on your luck. And losing against a totally luck-based enabling can really be frustrating.
|
 | #31 Feast (Base) Weighted Average: 30.25 / Median: 31.5 / Mode: 36 / Standard Deviation: 8.2 Highest Rank(s): #7 (1x), #12 (1x), #18 (1x) / Lowest Rank(s): #38 (1x), #40 (1x), #41 (1x)
Feast is the last card before a big gap. So its the best card of the bad third of the $4 cards. It has two outliers again with one in the Top 10 and very low mode.
Feast basically does nothing but being a one-shot balancing bad shuffle luck, especially at the start. If you really want a specific $5 card and have a 4/3 opening you can open with Feast and can be sure to get that $5 card soon. It also can be used with Throne Room and King's Court to gain multiple $5 cards.
|