(Start of Scythe strategy blog Here)
(Today, we once again have something completely different. GFLima, a BGG Scythe Forum regular and a frequent contributor to the comment section of this blog, has been kind enough to offer a comprehensive write-up on one of his own strategies. All that follows below are 100% his own words and thoughts. Please feel free to comment, as I am sure he is as eager to hear your thoughts as I am.
Thanks GFLima for your continued contribution to this blog!)
3-Worker Engineering Saxony
Whenever I played with Saxony I always produced five workers no matter what mat I had. I never questioned if this was the best thing to do. Five always felt safe, whereas 3 felt like too few, and the popularity hit of 8 felt too risky. Recently I realised that I was trying to play Saxony as if it were Rusviet (the faction I am most familiar with) but with a slightly more aggro vein. After all, they have the same initial terrain hexes, so why not just play the same moves in a different order? There are two reasons why this approach is flawed. The first is the fact that Saxony only starts with 1 power, whereas Rusviet starts with 3. Producing 5 workers with Saxony not only leads an extra production turn to go from 3 workers to 5, but also requires bolstering with no bottom row action early in the game. These are two small yet unnecessary expenditures of the commodity all factions want to save: time. Remember, Scythe is a race. The second reason, which is an extension of the first, is that power is one of Saxony's most valuable resources. It's something you do not want to spend needlessly if you want to engage in more than two combats. Three workers looks a lot more optimal for Saxony once you take this into account: you'll avoid the empty bolster action at the beginning, and you won't spend a resource that's highly important to you every time you produce. Besides, Saxony's access to wood isn't too terrible, they are only a riverwalk away, so it won't be too hard to build a mill and improve your production.
But still, can it be done? Can you build a decent engine with only 3 workers? I believe you can. I adapted the 3 worker strategy I had sketched for Engineering Nordic in one of my comments (see Case Study #3: Produce no Workers with Engineering Nords). Your engine will revolve around produce/upgrade and trade/deploy, then eventually bolster/build. This strategy makes a lot more sense for Saxony, which lacks and needs power, than it did for Nordic, which lacks combat cards. There are also other factors, and they would require a deeper analysis of Saxony's strengths and weaknesses, but that would deserve a post of its own.
Now I'll present a guide for the moves. I say 'guide' because there is at least one encounter, and, depending on what you choose, deviating slightly from the exact move order might make more sense. Hence this sequence of moves artificially disregards anything obtained in the encounter.
1. Move: Character to mountain or tundra, worker from mountain to village.
2. Trade: 2 oil.
3. Produce: 1 worker, 1 oil. Upgrade: move / upgrade. (You need this upgrade to get all your workers to the mountain without a mech)
4. Move: Character to village = encounter, 2 workers from village to mountain.
5. Trade: 2 oil.
6. Produce: 2 metal, 1 oil. Upgrade: bolster / deploy.
7. Trade: 1 metal, 1 wood. Deploy: Riverwalk on mountain.
8. Produce: 2 metal, 1 oil. Upgrade: produce, deploy.
9. Trade: 2 wood. Deploy: Speed on mountain.
10. Bolster: + 3 power. Build: Mill on tundra.
This is how your units are positioned:
Now you're ready to move your workers from the mountain to forest hexes, and take the second encounter at the same time, but exactly how you'll do it and what you'll do with your other move depends on a few factors. Here are a few options.
11A. Move: Character to forest encounter, mech with 2 workers to forest encounter, dropping one worker on the other forest hex, mech to tunnel on farm.
I usually only do this after checking if anyone can attack the lonesome mech there on the forest. Don't worry about the character and the mech there in the corner of the board, you'll build a tunnel.
11B. Move: Character to forest encounter, 2 mechs, one with 2 workers to forest, dropping one worker on the forest encounter hex.
This a more defensive move. I used it when the Automaszyna was on the tundra hex adjacent to the farm tunnel. I didn't want to be too close (since I hadn't bolstered enough), yet still within striking range.
Regarding what to take on each encounter, well, pretty much anything works, I would only avoid the third option since popularity can be an uphill struggle. I do recommend taking food whenever you can. It usually doesn't cost popularity and can be used to immediately enlist because it's the bottom row action under the move action. Although it's unlikely that you'll get the enlist star, unlocking the build bonus and gaining another 2 combat cards has worked extremely well for me. If you do this on the first encounter and both of your neighbouring factions are building you have a very good chance of hitting the third tier of popularity. And even if you don't get enough food to enlist straight away, since you're also upgrading, at some point enlisting will only cost 2 food, and you can save that recruit for later, when it will probably become clearer whether you need power, combat cards, or just a popularity bump.
I tested this strategy 6 times on Scythe Digital against three bots on the hard setting. Here are the results:
18 turns, $50 (popularity tier 1) - I decided to not care about popularity and try to end the game faster.
20 turns, $75 (popularity tier 2)
22 turns, $76 (popularity tier 2) - I suffered a mouse slip which cost me a turn...
21 turns, $78 (popularity tier 2) - lost to Crimea which was on the third tier of popularity and had 7 pairs of resources (???). I didn't really pay much attention to what was going on...
21 turns, $91 (popularity tier 3)
20 turns, $94 (popularity tier 3)
The two cases in which I was on the third popularity tier were due to unlocking the build bonus of the enlist on the first encounter. Now, I probably could have ended the games in fewer turns by taking advantage of the AI's rather predictable combat algorithm: if it deduces that it can't guarantee a win, it throws 1 or 2 power. Instead I played very cautiously, using enough power to make sure I'd win even if the opponent had the best combat cards and threw as much power as it could - and this meant that I had to bolster before engaging in the next combat.
I played two games against a single Rusviet Automaszyna (Automa level hard). On the first I had incredibly bad luck and lost with around $58. Basically, I had to attack it before it got its power track star, and when I did it threw 0 power and lost, only to bolster to 16 power on the next turn and subsequently attack me and win. The probability of something like this happening is very remote, roughly 10%. It is well known that if the Automaszyna reaches max power and wins a combat you can pretty much resign.
On the next game I wasn't so unlucky, and won with the exact same strategy, with $75 and probably around 21 turns. Yet I felt that this strategy wasn't optimal against the Automa. It could possibly be improved by not waiting for the speed mech to get to wood, bolstering sooner, getting your units closer to the centre of the board, since when the Automa is on phase I it will rarely attack, so you can stop it from spreading all over the place and also keep yourself within reach to strike it before it reaches 16 power.
I am generally happy with this strategy. It might be a bit slow (in contrast to Manhattan Jack's much faster strategies of case studies #1 and #2) but I feel that it suits my more defensive style of play. If combat isn't available you can just fall back on the bottom row action stars, and maybe a power track star, although it would have to be a very slow game.
Maybe in the future I'll compare this to a 5 worker strategy with the same faction and mat, and then maybe an 8 worker as well...
My thoughts on Scythe strategy. Also expanded to include updates on my foray into randomly generating more Scythe content...
- [+] Dice rolls