The Underlying Issues in Gwent

gerald witcher3 gwent



With many people discussing the current state of Gwent, I believe now is as good as time as explore some of the fundamental design issues in Gwent. The point of this post is not to highlight every issue Gwent has, nor to criticise specific cards, deck archetypes or keywords, but rather focusing on three non-exhaustive core, fundamental issues that Gwent has. When I do mention specific cards or decks, it is to simply as an example to illustrate my point, rather than to suggest that it needs to be changed specifically. Apologies for the wall of text, there's only so much I can do with formatting on Reddit. This post was in part inspired by this previous comment of mine.

The three underlying issues with Gwent are:

  • Low Point Range
  • Deck Polarization, Provisions, and the Gold/Bronze Disparity
  • Engine/Control/Point-Slam Triangle Imbalance

These are interconnected and overlapping. For example, low point ranges make it hard to balance engines in relation to control.

Low Point Range


A major issue with Gwent is that the average amount of points that an average card gives is too low, and this causes many issues with game design and balance. The best way to demonstrate this is by comparing it to the older, Beta Gwent. In Beta Gwent, a good bronze unit was worth roughly 12 points. By comparison, in current (Homecoming) Gwent, a good bronze unit is worth roughly 8 points (for 5 provisions). Balancing is more difficult in Homecoming, because each additional or subtraction of a point far more meaningful than in Beta. Adding 1 to a 6 point unit has a relatively greater effect than adding 1 to a 10 point unit.

A lower point range also means engines scale faster. A 4-point body that generates 1 point a turn exceeds its expected value faster than an 8-point body that generates 1 point a turn.

Expected Value of 1 point EngineBody PointsTurns to reach 100% EVTurns to reach 150% EV
8 (Homecoming)4 (5 on play)37
12 (Beta)8 (9 on play)39

This means with a low point range, engines become increasing more rewarding. The effect grows for more points the engine generates or combos with (such as Dol Blathanna Sentry). Having a higher points range, with bigger bodies attached to cards provides better soft cap for engines, as it is harder for them to excessively exceed their expected value when left unchecked. It also makes it harder for engines to boost themselves (or be boosted) out of removal range – there is a very narrow band of 4, 5 and 6 point damage removal in Homecoming Gwent. You cannot afford to chip down an engine by 2 points at a time when it boosts itself 1 per turn. The low points range homogenises card design too. Time for some Gwent trivia! What card am I describing: A unit with a 4 point body that generates (approx.) 1 point a turn for 5 provisions? Did you guess any one of 30ish cards? You are correct!

The low point range also has limiting factor on card and game design. A good example of this is the Beta Version of Triss: Butterflies, which used to be the counterpart of Yennefer: Conjurer. The ability used to be “At the start of your turn, boost all other lowest allies by 1”. This ability is a double-whammy for a low point range. Firstly, Low point range means it is easier to line up targets with the same points. 1-8 has less discrete numbers than 1-12. And, as we already discussed the more points an engine generates, the faster it exceeds its expected value, and Triss: Butterflies could easy hit 5 or more targets. The old version of Triss: Butterflies would be a huge pain to balance in the current incarnation of Gwent, if it was possible at all, which I believe explains its absence.

This similarly effects other cards. Cards like Regis and Geralt: Igni become substantially better the easier it is to line up points. The introduction of Initiative is arguably a work around the issue that low points range has caused. Other design areas are similarly restricted – a token cannot be less than 1 point. Creative weather effects like the old Skellige Storm (deal 3,2,1 damage to the first three units in a row respectively). And so on. The effect may seem minor, and in some regard, it is, but it is pervasive.

The solution is that Gwent needs powercreep, for bronzes in particular. Powercreep is a negative word in most CCGs, understandably so, as it’s often done thoughtlessly without considering how it affects the game, or to sell new content. However, in Gwent’s case it is needed, though it would likely take a long time to achieve. I do not know where the sweet spot for the points range in Gwent is. You do not want the points to be unnecessarily high and force players to do huge amounts of arithmetic. Previous experience from Beta Gwent suggests 12 for (5prov?) bronze, 15 for silver/low gold and 18 for an gold/high gold could be the sweet spot, or at least as a starting point. Alternatively a relative benchamark instead could be that Golds should be worth 150% of Bronze could be used instead. This does lead in to my next to topic…

Deck Polarization, Provisions, and the Gold/Bronze Disparity


The next major issue is the huge disparity between Gold and Bronze cards. The power disparity between the bronzes and golds is too high. In extreme gold cards are worth the equivalent of 3 or more bronzes cards. The most obvious example is Scenarios, which literally play for 3 bronzes + synergy. Though, this applies to most high-end golds which are generally worth 20-30 points, or equivalent to 3 or more bronzes, if not literally. The result is that this polarises the game around golds, and consequently increases the impact of draw RNG. It also makes games incredibly swingy – does your gold card beat my gold card? The value of tempo and card advantage is also diminished, when one of your golds can play for 3 of your opponent’s bronzes. The strategy for bronzes largely revolves around not playing them or burning them round 1. With a few exceptions, you are never in a situation where you want to play bronzes, but rather you are forced to.

Part of the problem is how provisions scale. The formula in the expected value (good) cards currently seems to be “Power = Provision Cost + 3, minor conditional”. Examples: Aen Elle Conqueror, 7 for 4. Bear Witcher, 8 for 5, Berengar, 9 for 6, and so on. However, this formula breaks down as you approach higher provision cards. While it’s harder to evaluate the expected value of (good) high end golds, due to more complex mechanics, utility and synergies, it does seem to be above “Power = Prov + 3”. Some examples include Gezras – a very conservative 2 procs of 5 already puts Gezras equal to the formula (5 + 5 + 5) 15 for 12 prov. Harald the Cripple is usually 14 on play, plus 1 more for every Warrior, for 12 prov. Siege is conservatively a (8 + 7 + 6 = 21) for 14 prov. As the provision cost increases, so does the provision efficiency. This means players are incentivised to play as many high end golds as possible and polarise their decks with terrible 4p bronzes they want to avoid playing (or are tech cards) and 10+ golds. It’s a no brainer why decks like Lippy or Double Ball became prevalent.

Some people may object and say “hang on, 7/4 = 1.75, and 21/14 = 1.5. Low provisions are more efficient!” You need to keep in mind that the estimates for high end golds are conservative, and can and often do exceed this number. More importantly, you only play 16 turns (in most games). Your goal is to produce as many points as possible in those 16 turns. To achieve this goal, it is significantly more effective to dump those points into high provision cards, rather than go for a more balanced deck building approach. High provision cards should have significantly decreased provision efficiency and move away from raw points which cards like Scenario, Evolving and the new Leaders cards are, and focus on them becoming more tech cards and bronze dependent.

To compare Homecoming Gwent to Beta Gwent again, there has been a profound shift in design for the role of bronzes and golds. In Beta Gwent, bronzes were your point generators and win conditions, while silvers and golds most acting as tech cards, enablers or tutors. All the iconic decks of Beta Gwent rely on you playing your bronzes. SK Greatswords, Queensguard, NG Alchemy, Nekker Consume, Henselt Machines and so on. In Homecoming Gwent, this dynamic is largely reversed. Recent meta deck, like Elves Movement, NR Witchers, SK Warriors, Viy, Lippy etc. are all completely dependent on playing your high-end golds. Of course, there are exception in both versions of Gwent, but the general design direction holds true. Golds in Beta were only worth 50% more than bronzes (18/12) compared to 200% (21/7) or more in Homecoming.

The solution is that the power difference between bronzes and golds in Gwent need to be extremely reduced. Ideally this would match with the previous issue by primarily powercreeping bronzes, but would require more extensive rebalancing and reworking of all cards. Cards like Scenarios could not exist in their current form in this rebalanced Gwent, where they literally play three bronzes. This would be no easy or quick task, and potentially impossible with the current state of Gwent.

Engine/Control/Pointslam Triangle Imbalance


This issue is in part a consequence of the previous two issues. The Gwent Engine/Control/Pointslam Triangle (“Gameplay Triangle”) is Gwent’s version of the rock-paper-scissors system found in many CCGs and other types of games. In many CCGs this takes the form of the Aggro/Midrange/Control where each one will beat the other, though the direction may depend on the specific CCG, and the presence of Combo decks as a 4th general archetype.

Gwent substantially differs mechanically from the MTG family of CCGs, but it too has its own gameplay triangle in Engine, Control and Pointslam. The Pointslam strategy is what the name implies – it is a strategy the revolves around playing cards that generate immediate, raw, points, typically, but not always, in the form of cards with big bodies. It strongly tied to the concept of “tempo”, or generating a large amount of points in a minimum about of turns such that your opponent cannot catch up in one card. Engines are the opposite – it is a strategy the revolves around playing cards that generate points over several turns. Engine may be a bit misleading and it could also be called “Value”, because it may include cards that are not strictly “engines” but still need multiple turns or other cards to generate points, such as Yennefer of Vengerberg. Lastly, Control is a strategy that revolves around denying your opponent from generating points to gain an advantage, rather than developing your own points. These strategies are not mutually exclusive, and any given deck may contain elements of all three to different extent. Ideally, you would want to incentivise players to have elements of all three to promote dynamic and interactive games.

The three strategies theoretically have a rock-paper-scissors relationship. Engines beats Pointslam, as it generates more value than Pointslam in the long run. Control beats Engine, as prevents Engines from getting full value. Pointslam beats Control, as it produces more immediate points than Control can remove.

To use an example with arbitrary points:

Engine card is worth 5 points, and generates 1 per turn. The Control card is worth 7 points, meaning when it removes to the Engine, it gains 1-2 net points. The Pointslam card is 9 points, meaning it beats Control by 2 points, but loses to Engine over a few rounds as it generates more points than the Pointslam.

The issue in Gwent is the balance between the three archetypes are imbalanced, and there is currently strong incentive to polarize your deck as either a greedy engine/value deck, or a hypercontrol deck, including “unitless”.

There are many contributing factors, including the two major above points.

  • Engines are too efficient
  • Most points come from golds, meaning you do not have to generate points from bronzes in control decks.
  • Tempo is less important in Homecoming (in part due to 10 card limit).
  • Most (bronze) control options do not generate net gain of points when played.
  • Pointslam, outside of a handful of gold cards, is awful. Bronze pointslam cards are particularly bad.

The last two are worth expanding upon.

Control tools rarely produce a net point gain, particularly bronze control cards. For example, using an Assassinate on An Craite Longship is a neutral point exchange. But perhaps even more commonly control cards are a net point loss, if the point from the engine is not attached to the body such as Temerian Drummer or Cat Witcher. This means there is little incentive to splash (add a small amount of) bronze control. Instead of removing your opponent’s engine and go negative on points, it is often more efficient to try and develop your own engines instead. So instead control only gains value from complete denial of your opponent’s strategy in hypercontrol. This includes unitless decks, who get further value by not only denying your opponent developing their board, but also denying them the ability to interact with your board.

Pointslam as an overall strategy is effectively non-existent. It exists only in a very limited degree in some gold cards, such as Yrghen, Cerys and Lippy. Bronze “pointslam” cards are incredibly bad. Most are conditional 7 for 5, such as Savage Bear, or Ard Feainn Tortoise. The only really viable pointslam bronzes happen to be from the recent, powercreeping expansion, such as Bear Witcher and Viper Witcher Mentor. A major reason why bronze pointslam is so bad is because there’s very little utility in them. They do not generate enough tempo on their own to justify running them over engine or control. Why run a conditional 7 or 8 for 5, when you can run a 5 for 5 with a lock, purify or some other utility?

As the imbalance of Gwent’s Gameplay Triangle is partly downstream from the other issues, solutions to them will help this issue also. However some additional solutions include reducing the conditionality of bronze pointslam (it undermines the whole point of pointslam) in addition to powercreep. Control needs a way to generate net positive points against engines to make splashing control viable, in Open Beta this was achieved through tutors. One good example of how this has already been achieved to some extent in Gwent is the Symbiosis keyword, however this is still conditional on LA or other cards. Minor damage/removal is also heavily overvalued. Cards such as Aedirnian Mauler, Dol Blathanna Archer, should be comparable in value to other cards at their provision cost. Minor damage cards encourage interactivity and reward more thoughtful sequencing.


These are not the only issues with Gwent’s current state and design, but I do believe they need to be addressed for the long-term health of the game.


leave a comment

Your email address will not be published. Required fields are marked *