Push the Lane: Loot in a Strategy Game?

Since Push the Lane entered this latest phase back in mid-2017 (basically after the failed Kickstarter version, which was much more puzzle-game-like), it has become much more videogamey. By that, I mean, it has focused a lot more on fighting, monsters, items, special abilities, moving around a big map and such. I have been thinking of it more like “a Rogue-like DotA” recently; a turn-based, single player League of Legends.

With that thought, I always kind of had it in the corner of my mind somewhere that it would be pretty cool if the game had “loot” somehow. My general feeling and belief about loot has been, for years, that it has really no place in strategy games. But maybe there’s a way? First, let’s define the term.

What is loot?

I think most of the time the word “loot” is used, it refers to randomly dropping items. For me, the classic version of “loot” is item drops in Diablo, or a Rogue-like. More recently, it’s popular to have “loot crates” in games like Overwatch, which give the player some random metagame items, such as skins.

I think Push the Lane will probably have both. The second kind – the Overwatch kind – is pretty easy to implement into a strategy game without damaging it too much, as long as you make sure to only let the player unlock aesthetic stuff or sometimes side-grades. But the first kind, the Diablo kind, is a lot harder to implement into strategy games.

In strategy games

The quickest way I can express why it’s hard to put loot of this kind into strategy games is to say that it’s simply unfair. In a two-player game, it’s pretty obvious that I shouldn’t win a match against you because I happened to get better item drops than you did.

In a single player game, it might be less obvious why it’s a problem that I won match 1 because of a crazy good item drop and lost match 2 because of the lack of good drops. You might feel like, well, so what? Who cares? Worst case scenario, you’re just cheating the computer or something, right?

The problem might not be too much of a problem while you’re climbing the ranks (see this article to read about Single Player Elo systems). But once you get to your skill level’s appropriate rank – once your win rate has started to level out to somewhere near 50% – that’s when stuff like random drops starts to become a problem. When the game is really hard, when you’re striving to win in a very difficult atmosphere, stuff like random drops can really throw that off.

This isn’t as much in something like a high-score system Rogue-like game, because there isn’t a binary win/loss condition (or at least, that’s not usually what players are playing for in such games). In a Rogue-like, maybe you get a bunch more loot one game, and so you end up with a bigger score. It’s less significant, because it’s not really clear what you are trying to do goal-wise anyway. With a win/loss based game like PTL, you’re clearly going for a win, so if you get the win because of item drops, that’s a problem.

The inclination, then, is to basically make them less random in order to make it more “fair” (or to use a better term, make it more uniform). That means a smaller variety of what can drop, and less surprise about when it drops. So a hard other-end-of-the spectrum would be to make monsters always drop the same item, or maybe a very small pool of items that don’t vary much in value at all.

That is almost certainly something that would work in Push the Lane – but would it be loot? Also, do I really care if it’s loot or not? And if so, why?

Why I care

One of the things that has been missing from the game so far is new, meaningful, big and chunky information coming into the game after the start. Most of the randomness in current PTL happens at the start of the match: which zones you have, where the doodads are, and which items are available in the shop (as well as item costs). In terms of mid-match randomness, the only things that really happen are when the bosses appear (which is significant) and when monsters and orbs appear (a lot less significant). We need more.

And further, we need stuff that will really surprise players, in both a positive and negative way. Right now most of the surprises that happen are negative – oh, that monsters spawned right where I didn’t want him to, or oh, that boss spawned right when I didn’t want him to. There isn’t much of a “yay, this cool thing happened that I didn’t expect”.

I think that with strategy games, you want players to be able to predict stuff enough that they have a strategy, but it needs to be a flexible strategy. That course of action they’re going to take needs to be able to bend, and the player needs to be able to make it work. The player needs to be able to not just handle bad surprises, but also know how to take advantage when they get a good surprise.

I also think that Push the Lane being kind of a strange game, we could use as many hooks as possible to pull players in. I think one of the ways that Rogue-likes are able to do that is by using this sort of operant conditioning random Skinner box thing. Players play, to some extent, hoping that they will get some crazy payout. Even Civilization kind of uses this to some extent: when I start new matches, part of the excitement is in grabbing all the “tribal villages” (very offensive) – maybe this match I’ll get a free Scout super early on, or some clutch technology out of it.


So what I’ve been struggling with is how to create a system that gives the player, overall, somewhat uniform value, while still feeling random enough to feel like loot. I’ve gone through a lot of different formulas and weird ways of doing it.

My plan now is to let it be a little bit wild, within some parameters, and hope that because the matches are pretty long and because there are multiple axes on which random events are happening, there will be enough going on so that the match isn’t won or lost based on what loot you have.

I think something that I’m grappling with is that I’m no longer making a tactics game, like I did with Auro. I’m now making a strategy game, and it’s a decently big strategy game. There’s a lot going on. So while getting some random item in Auro could probably have thrown the whole match out of whack, in Push the Lane I think it might get somewhat absorbed.

Also helping this is the fact that you don’t really want more than a certain amount of most stats, and you also kind of want some of all of the stats. You can sell items (at a loss), so there might be a few situations where you do get a lot of really good items, but in order to cover some bases you’re missing (like, let’s say you get a few really good offensive items, but no defensive ones), you might have to sell something off in order to get some defense.

Here’s some details: there are a few treasure chests around the map, including some in the Forest. These each contain some random items (low to medium value). Also, most Guardian monsters (unlocked by capturing capture points on the lanes) drop a low to medium value item. Finally, the two “lane-end chests” at the end of the left and right lanes drop a medium to high value item.

I have some math and some averages that tell me how much general value these should drop. But beyond that, I also probably will put in some “caps” at some point: if the player has already achieved, say, 10,000 gold worth of value this match, the game can go into a “very lame loot mode” where things either drop nothing or very little. We can do the same on the other end – if the game hasn’t given you much in the first 200 turns, we can make sure the next thing it drops is pretty good.

It’s possible that even without measures like that, the game could work out and be okay. But I think if we have those in place, it will ensure that stuff can only get so out of whack. And there’s so much the player can do in terms of strategy and just general input that I think their skill should be able to overcome whatever comes their way. That’s another problem, which I have to remember, is that a lot of these unfair loot games (Rogue-likes come to mind), the skill cap is actually pretty low. Other than knowing what everything does, there isn’t a lot of strategic skill you can gain at playing a Rogue-like game (even if there’s arguably a good amount of tactical skill there).

Anyway, it’s back to the programming mines for me now. Hope you enjoyed this, and please let me know if you have any thoughts on what I’m doing with loot in this game. See you soon!

If you enjoyed this article, consider becoming my Patron at Patreon.com!

  • Rob Seater

    A few options come to mind for how to use random loot in a strategy game:

    (a) Generate the loot up front, and either let the player see it right away or let them scout it out over the course of the game, prior to committing to what loot they will discover. Several good euro games use randomized secret goals this way, where part of the game is determining the value of information vs. the pursuit of goals informed by that information. If certain locations on the board (that you can fight over to, spend resources on, or build on top of) reveal what loot is at the ends of the lane, then how to adjust your strategy based on that information becomes more skill than luck.

    (b) Allow all loot to have some default value. E.g. if every loot is an item I can use, but every item use consumes a piece of loot, then most loot will just be fuel. A randomly bad drop isn’t so bad, because most loot is just fuel. E.g. Race for the Galaxy and several card-driven strategy game use this approach to smooth out random draws while still offering a lot of decision and skill. That requires that the player receive a lot of loot (or draw a lot of cards) for the luck to average out and the number of viable item uses to have a meaningful impact on the game. It also consumes a lot of the player’s time (and the complexity budget for the game) so you had better make the loot interesting and powerful enough to be worth the player stopping to study their inventory. In a game like RftG, that’s pretty much the whole game.

    (c) If you are thinking about just a few valuable pieces of loot (e.g. at lane ends or from major boss-like adversaries), then another option is to make them all very powerful but require adjusting your strategy to match them. Perhaps when you get a piece of loot, all your equipment is immediately liquidated for full value, you get a random bonus/power/whatever, and you can re-spend your money to adapt your strategy to your new item. A novice may get an item they don’t know how to use, but you can’t really complain that you got something that didn’t fit your strategy, since the game just gave you a chance to change your strategy. Some very deep strategy games use that kind of randomness to create a lot of tactical choices rather than tightly controlled strategic choices. It adds replace value, since you can’t predict want item (or combination of items) you’ll get and thus can’t just stick to the same strategy every time. However, it can harm the player’s sense of identity and strategic expression to have a strategy they were building towards taken away from them mid-game — some players will find that interesting and some will find it unsatisfying. If the items are fun and varied and toy-like enough (weird but powerful abilities, not simple stat boosts), then the player will feel excited to try to make the new thing work, rather than disappointed that their old toys went away.

  • The tricky thing with powerful and interesting loot in singleplayer games is that particular pieces of loot tend to combo in certain, very powerful ways. At some point you stumble on a particular, very powerful strategy, exploit it every time and start to get frustrated when you don’t get loot required for this strategy. The example would be Hearthstone’s new rogue-like-like Dungeon Run mode, where after getting your first and second combinations of cards and artifact you’re often tempted to restart the run for a better one. It can be argued that “playing wrong” is player’s problem, but I’m convinced that psychological pitfalls are always fault of the system, not the player.
    One of the ways to mitigate this is to make sure you have multiple ways to get crucial items, including deterministic ways. Spelunky does this, ’cause you can get jetpack in the cave zone – but still, I get very frustrated and tempted to restart if I don’t get jetpack in jungle zone.

    I second Race for the Galaxy, which solves this problem nicely. I think it’s one of the most skill-based games I’ve played, despite it’s very random card game nature. One of the reasons for that is that most of the cards have multiple powers, each of which can actually be used as a cornerstone in several builds. So when getting a single card, you actually get an opening into a half dozen of different strategies – and you cycle lots of cards throughout the game.

    I suppose, for PTL system I’d use something like this:
    – each game has a hidden pool of 10 supercool items (chosen from several dozens that could actually exist).
    – whenever you get big loot, you get to chose one from three (which are drawn from the pool of 10).
    – whatever is not chosen, returns to the pool, but gets a permanent bonus for the next time it is drawn.
    – dunno if psychological effect for small and medium loot is significant enough to even include it.

  • Jereshroom

    That sounds like it would create too many unintuitive, luck-based strategies. Like, I would want to start by picking the loot that gets the worst not-picked upgrade, and then hope I get the right loot multiple times so that when I finally pick it it’s stronger than usual.
    Choosing one of three seems good, but having cards upgrade when unchosen is too abusable. (It seems like a mechanic for when the dev doesn’t have the ability to properly balance anything.)

  • Jake Forbes

    I agree with a lot of the thoughts offered by Rob and Random. The draft variant of “pick 1 of 3” model is potentially a great fit because for a strategy gamer, evaluating and choosing is intrinsically rewarding, potentially more than the loot itself. RftG’s does a phenomenal job balancing cost and income – more so than any CCG I’ve seen. The more recent game Terraforming Mars satisfies in a very similar way. Both games benefit from a huge diversity of cards and a high circulation to allow for adaptable strategies. I’ve brought up Dream Quest before and will do so again – it is a rogue-like based on card drafting and it adds a tension inspired by Dominion where deck efficiency is key. There are maybe 25 opportunities to draft, give or take — 10 mandatory, the rest opt-in. It manages to find a sweet spot between a mathy efficiency puzzle like Dominion and a delightful but swingy draft like Hearthstone’s dungeon run.

    The more appealing and dynamic you make the loot, the more likely it is that you’ll have a subset (maybe even a majority) of players who prioritize chasing that epic win from the perfect combo more than they do their win rate. This might be the “wrong” way to be a good strategy game player, but that’s okay because it broadens the paths to fun in your game. Allowing some totally broken combos to come together very rarely isn’t so bad as for novice players, it lets them feel smart about finding synergies, while veteran players know the odds of the perfect combo are so low that chasing it is a sucker’s bet when you’re focused on your ranking.

  • Really good thoughts. There’s a lot of levers here and a lot of ways we can do it. I think one thing I’m counting on is that in this game, you can do a lot of stuff. You have a lot of options and there aren’t like 50 turns or something, but like 500-800 turns. There’s a big ass map and just a ton going on. So I am hoping that even if there are some swings in stuff, you will be able to make stuff work. Some factors working in my favor are: you can sell unwanted items and buy (random, but still visible for the whole match) items – that helps. But also, you generally will need all of the stats, I think pretty much every game.