Stim is truly a wonderful ability. It is arguably the most powerful upgrade in the game, its completion signals the switch from Terran playing defensively to Terran taking the offensive.
New players often don't understand its power, not understanding just how much strength is gained from it, and being unwilling to spend the health it costs. Once they get over that hurdle and embrace its potential, they often make the mistake of over-stimming. Some players even don't seem to understand that it only works once, and lose all their health stimming multiple times at once.
But what if there were a way to benefit from multiple stims? As it turns out, there is. It's not easy though.
To do a double-stim, you need to shift queue exactly 31 move commands to the exact same spot, queue a stim command, and then manually issue a stim command on the same frame they reach their destination. I suggest using rapid fire for the queueing of the move commands. It's pretty hard to time it correctly, but it's kind of similar to queuing up 31 marines on a bunch of barracks, just without the visual indicator. Also, make sure you don't have any medivacs selected when you do this, because it'll screw with the marine pathing so they'll run in circles around the destination.
Uses In my last post on oracles, I made the mistake of trying to predict the potential uses for the behavior before really testing anything out. Thus I shall repeat myself and make that mistake again. It is not possible to triple-stim, but I doubt it would be a valuable trick even if it were, because the health loss is too great. As it turns out, if you have medivacs present, a double-stim is indeed a bit better than a single stim in a straight up fight. However, by far the greater benefit is if you have something to tank for your marines, such as scv's.
What situations will this be useful in actual games, accounting for strategic considerations? I can't say, as I'm not a terran player. Maybe if players get very consistent with this it would unlock a 2 base scv/marine all in timing. Maybe it will just be reserved for the rare situation in which it's useful. Or maybe it will be deemed to difficult and too high risk/low reward to ever employ in a real game.
Explanation For The Nerds
Warning: Technical Expertise Required
The game only buffers 32 queued commands for a given unit. By queuing up 31 move commands before to the stim, we get the stim command to be at the end of at the end of the buffer. There is an off by one error in the engine that allows us to abuse a race condition to push a second stim behavior onto the unit before the relevant check. By queuing all the move commands on the same spot, they will all be removed from the buffer on the same frame. Stim is a transient ability, which means that issuing a stim command won't cancel the queue. This allows us to issue a stim command at the start of the frame the marines empty the buffer. The stim behavior existence check happens before a stim order is given, but the stim behavior application happens after the order is executed. Normally this is fine and results in the expected behavior. However, Due to the error off by one error mentioned earlier, the final order in the action queue buffer gets executed in between the check and the execution of the manual command. This causes the correct order of execution to be bypassed. Surprisingly, this is possible even though Starcraft is a deterministic and synchronous game that runs in lockstep. Thus there is no nondeterminism requiresd, simply an abuse of an off by one error that capitalizes on a flaw in the buffer code, not to cause a buffer overflow, which would allow for far more deleterious behavior than a humble double-stim, but rather to inject (not SQL, don't worry) the execution of a command where it doesn't belong. Thus the integrity of the checks-effects code flow is compromised and the usual restriction of a single stack at a time of the stim behavior is circumvented allowing for the clever user to benefit from a double stim. Unlike the situation with the oracle, in which I discovered a potentially (later confirmed to be definitely) unintended behavior, which I still claim should be called a feature rather than a bug, and which has seemingly been accepted as such as indicated by the retraction of the supposed "fix", I do recognize this trick as being the rather blatant abuse of what can only be considered a bug. There are no two ways about it, this is definitely not intended, and I can't see a reasonable argument being made for its intention. However, should it be removed on those charges? We could consider similar cases from other games, such as Smash, or even Brood War, in which blatant abuse of obvious bugs is commonplace and even mandatory for high level competition. The argument could well be made that a game such as Melee would be far inferior if all of those bugs were to be fixed. Whether the presence of this specific bug results in interesting or degenerate behavior is hard to predict and maybe is best left to the community to observe and judge. For now maybe adhere to the classic "let the meta settle", and if it settles in an undesirable way, then perhaps intervene. Although I doubt Blizzard would take that approach. Anyway, the summary I guess is that even though there is no nondeterminism requiresd, simply an abuse of an off by one error that capitalizes on a flaw in the buffer code, not to cause a buffer overflow, which would allow for far more deleterious behavior than a humble double-stim, but rather to inject (not SQL, don't worry) the execution of a command where it doesn't belong. Thus the integrity of the checks-effects code flow is compromised and the usual restriction of a single stack at a time of the stim behavior is circumvented allowing for the clever user to benefit from a double stim.
Source: Original link
© Post "Stim – The New Beam Weapon" for game StarCraft.
Top 10 Most Anticipated Video Games of 2020
2020 will have something to satisfy classic and modern gamers alike. To be eligible for the list, the game must be confirmed for 2020, or there should be good reason to expect its release in that year. Therefore, upcoming games with a mere announcement and no discernible release date will not be included.
Top 15 NEW Games of 2020 [FIRST HALF]
2020 has a ton to look forward to...in the video gaming world. Here are fifteen games we're looking forward to in the first half of 2020.