Why Did a Missing Horse Skin Make Me Actually Sad? A Live Octalysis Autopsy of Heroes of the Storm

Halfway into a Heroes of the Storm match, I realized my horse was basic.

Not my champion. Not my weapon. My mount. The animated blob of pixels I ride for six seconds between respawns. The cool skin I’d paid for wasn’t equipped, and instead of shrugging and getting on with the match, I heard myself say out loud: “Oh no, my horse is basic. I don’t have the cool horse skin. Nooo. I’m so sad.”

Welcome to a live Heroes of the Storm Octalysis.

And I was sad. About a virtual horse. That changes nothing about the game. That nobody on the opposing team could even see during our fight.

Here’s the twist: I teach this stuff for a living. I’ve built a whole framework around why human beings respond to design elements that, logically, should mean nothing. And I still felt it. The cool version of something I don’t actually need, withheld, produced a genuine emotional dip — the same kind of small internal collapse you feel when you realize you grabbed the wrong coffee order from the counter.

That moment is the reason this post exists. If you want to understand why cosmetic monetization prints money, why streamers pick the flashy ability over the strong one, why losing players still tip each other, why veterans want the generic currency more than the legendary prize, and why an “XP-only” lane can turn into the most stressful job in the game — you don’t need a survey. You need to watch a behavioral designer narrate his own gameplay in real time and tell you, with a straight face, that yes, the irrational thing really happens. Even to the person who invented the framework.

Speed Run Notes

• Cosmetic ownership triggers Core Drive 4 (CD4) whether or not the player believes in it. I knew the horse skin was purely visual. I felt the loss anyway. That’s the feature, not the bug.
• When an audience is watching, players switch from “best” to “coolest.” Core Drive 5 (Social Influence) amplifies Core Drive 4 (Ownership) at the direct expense of Core Drive 2 (Development & Accomplishment). Streamers aren’t playing the same game as ranked climbers.
• Post-match social voting is a loss-softening mechanic. In a game where half the players lose every round, letting teammates up-vote each other answers the unspoken question, “was my effort worthless?”, with “no.” That’s Core Drive 5 doing retention work.
• Advanced players prefer the generic currency (shards) over the specific prize (legendary skins). Duplicates turn from a “bad” outcome into useful progress. Currency converts Core Drive 7 randomness into Core Drive 3 meaningful choice.
• Shared team leveling fuses Core Drive 1 (Epic Meaning, because I soak the lane for the team) with Core Drive 8 (Loss Avoidance, because if I don’t, everyone falls behind). That’s why HotS lane-soaking feels more like a duty than an optimization.
• The “quarter-kill” character (Murky) teaches a business lesson in disguise: not every target is worth the same kill. The designer decides what’s worth chasing. You get to do that too, in pricing, in outreach, in hiring.

Why I’m the One Writing This

I’m Yu-kai Chou, creator of the Octalysis Framework, the eight-Core-Drive behavioral design system that’s now applied to products, public programs, and game experiences reaching hundreds of millions of people. I’ve advised MrBeast, LEGO, Microsoft, Porsche, Coca-Cola, and governments including Ukraine.

I’m also, as the opening paragraphs confess, a person who plays a lot of Heroes of the Storm. I mean it when I say the horse-skin moment was real. This post is built from an unedited walkthrough of my own play session. The emotional reactions are primary-source data, not a case study I read about.

That matters for this topic, because cosmetic psychology is famously slippery in surveys: players under-report the strength of their attachment when asked directly, and over-report their rationality. You only catch it by watching someone feel it in real time. The value of me being the one writing this is that I’m willing to catch myself.

The framework underneath every observation below is the Octalysis Framework, explained in full in my book Actionable Gamification. If any of the Core Drive references below feel underexplained, those are the two places to go deeper.

The Horse Skin Moment: CD4 Without Logic

Start with the cleanest evidence. I opened a Heroes of the Storm match, picked my hero, rode my horse out of the base, and realized the skin I’d paid real money for wasn’t equipped. The horse was the default brown mount every new player starts with.

My internal reaction, said aloud on stream: “Oh no, my horse is basic. I don’t have the cool horse skin. Nooo. I’m so sad. See, this is funny. It doesn’t matter, right? It doesn’t matter. But I could be cool and I’m not cool.”

That last line is the entire lesson in twelve words. I could be cool, and I’m not cool. The gap between the self I’d paid to construct and the self that showed up on screen felt like a small loss. Not a huge one. Real.

Core Drive 4, Ownership & Possession, is the Core Drive most people misunderstand. They think it’s about the thing you own. It’s not. It’s about the self the thing extends. My horse skin never affected gameplay. It affected the identity I was wearing while I played. When the identity downgraded without my consent, a piece of me went with it.

Later in the same session, I teleported back to base, watched the cool animation I’d paid for, and said, “This is what I paid for. Looking cool when I go home.” Notice the phrasing. Not “this is the animation I bought.” This is what I paid for. The purchase wasn’t an animation. What I bought was the moment of feeling cool. The animation was just the delivery vehicle.

This is why “cosmetic only” monetization is the quietest money-printer in game design, and why live-service games that try to sell only functional upgrades keep hitting a ceiling. Functional upgrades trigger CD2 (accomplishment) and are easy to calibrate into “pay-to-win” complaints. Cosmetic upgrades trigger CD4, which has no obvious ceiling, because identity has no obvious ceiling. You can always be a slightly cooler version of yourself.

If you design products outside of games, the lesson is blunt. Stop arguing with users about whether your premium tier “really” delivers more value. If the premium tier makes them feel like a different kind of person when they use your product, a sharper, cooler, more serious version, they will pay for it, and they will not be able to explain why on a survey. I covered the self-extension logic further in The Grown-Up Lock, which looks at what happens when ownership and usage diverge. The horse skin incident is the same engine, running in the opposite direction: usage without the desired ownership feels worse than no usage at all.

Style Over Strategy: When CD5 Hijacks CD2

Later in the match, I had to pick my Heroic ability, the ultimate move that defines how a character fights at high levels. There were two legitimate choices. One was, by any analytical metric, the stronger pick for the team composition we were running. The other was called Judgment, and it had a much flashier animation.

I picked Judgment. Out loud, I said: “The reason I’m choosing Judgment is not because it’s the best heroic. It’s because it looks cooler. And style is important. Especially when I’m streaming.”

Those last three words are the whole design insight. Especially when I’m streaming.

Alone, I probably pick the stronger ability. Climbing in ranked, I definitely pick the stronger ability. But the moment an audience enters the system, the math changes. CD5, Social Influence & Relatedness, arrives at the table and starts shifting the weight of every choice. Looking cool in front of strangers climbs in value. Marginally improving my win rate drops. A trade I’d never accept alone becomes obvious in public.

This is not a character flaw. It’s a predictable Core Drive interaction. When CD5 is active, CD4 choices get louder and CD2 choices get quieter. Designers who understand this ship different products for audiences that will be watched (TikTok, Twitch, multiplayer party games) than for audiences who play alone (single-player RPGs, productivity tools). The first group will sacrifice performance for expression. The second group won’t.

I’ve watched this exact pattern destroy B2B SaaS launches. A vendor builds a tool that’s objectively more efficient than the one the team already uses, ships it, and watches adoption die. Why? Because the new tool makes the user look like a different kind of professional in Slack screenshots, and the user doesn’t want that identity. The old, clunkier tool let them look competent and ordinary. The new tool, no matter how powerful, made them look like they were trying too hard. CD5 hijacked CD2, and the procurement team never knew.

If you’re launching a product that will be seen by the user’s peers while they use it, audit the visible artifacts before the features. Screenshots, notifications, the little status badge next to their name: those are identity claims. You can out-perform every competitor on efficiency and still lose on identity. It’s the same mechanic that makes me pick Judgment.

Post-Loss Voting: Designing Dignity Into Defeat

We lost the match.

In most competitive games, a loss is a clean negative. You see a “Defeat” screen, lose some ranking points, maybe get a stats breakdown, and sit with the disappointment. Heroes of the Storm does something different. After the loss, the game shows every player’s performance stats, and then invites the entire lobby — both teams, all ten players — to vote on who played well.

My reaction on stream: “They get to see, make people feel good. So you get to vote. These are what we call the social pros. Just, hey, thumbs up. Makes people feel great. Even if you lost, hey, you still got some kind of recognition award.”

Notice what the voting system is actually doing. It’s not grading performance; the stats screen already did that. It’s answering a specific unspoken question the losing player is asking themselves: “Was my effort worthless?” The voting system answers, “No. Someone noticed.”

This is CD5 (Social Influence) doing retention work. In a game where exactly 50% of all players lose every single match by structural design, the default emotional arc of the playerbase would be a slow grind toward churn. You can’t avoid losses — half the content on any given night is a loss. So you have to design the loss so it doesn’t produce the feeling that leads to quitting.

Most designers approach this with system-generated consolation: participation XP, “good effort” badges, end-of-match gifts. Those work, but they’re obvious. Players know the game is handing out the trophy. The HotS voting system routes the recognition through other humans instead. A thumbs-up from a teammate who just watched you play carries weight a system-generated token never will. The game’s only job is to provide the surface where the humans can tip each other.

The business parallel writes itself. Every customer interaction has a losing side. The deal that didn’t close, the support ticket that ended in escalation, the feature request that got declined. Most companies treat those as transactions to survive. The best ones build peer-recognition surfaces that let the losing side walk away with their dignity intact: a personal email from a senior exec, a shout-out in a community Slack, a named contribution in a changelog. The losing interaction becomes a retention event, not a churn event.

For the deeper mechanics of why competitive games produce so much raw emotional volatility in the first place, and why designers have to build shock absorbers like the voting system, I wrote the system-level version in Why Games Turn Toxic.

The Duplicate Paradox: Shards Over Skins

Later, I opened a legendary loot chest, the highest-rarity reward in the game, and got a legendary skin I already owned. Duplicate. The game converted it into 400 shards, the generic currency you can spend on any skin you want.

A newcomer to the game would read that as a bad outcome. You opened the top-tier chest, didn’t get a new top-tier item, got some currency instead. The rarity icon says legendary; the outcome feels common.

My actual reaction: “I got a duplicate, right? So it’s a legendary skin, but it’s duplicate. So I got 400 shards. Which actually is very welcome. I like the shards. Shards are valuable because it gets me what I want. Versus just getting a legendary that I don’t care about.”

Read that last sentence twice. I’d rather have the generic currency than the specific legendary prize, because the generic currency lets me choose, and the specific prize might be something I don’t want.

This is the currency preference inversion, and it is one of the most underrated insights in reward design. Early in any game, players want the specific rewards: the named legendary, the exclusive skin, the themed mount. They haven’t formed preferences yet, so any legendary is a good legendary. Late in the game, after preferences have crystallized, the specific reward becomes a gamble. Will the drop align with what I actually want? Usually no. The probability of a legendary matching my taste is low. The probability that shards let me buy exactly what I want is one hundred percent.

What’s happening in Core Drive terms: a direct random legendary drop is CD7 (Unpredictability) wrapped around CD4 (Ownership), with a nasty CD8 risk (Loss Avoidance, the dread of wasting a chest on something you don’t want). Converting the duplicate to shards collapses the CD7 and CD8 into CD3 (Empowerment of Creativity & Feedback): meaningful choice. The player stops being a passive recipient and becomes an active chooser.

If you design any reward system, whether loyalty points, referral incentives, or holiday gifts for employees, stop shipping specific prizes by default and ship generic currency backed by a rich catalog. A $50 random gift card to a store the recipient doesn’t shop at is worse than 50 “points” that buy anything in a catalog of 100 options. The $50 gift card is CD7 roulette. The 50 points are CD3 agency. Every advanced player on earth has already learned this lesson in a video game. Your customers have learned it too — they’re just waiting for you to catch up. Reward Design: Why Most Rewards Backfire goes deeper on the full taxonomy of which rewards help, which hurt, and why the defaults are wrong.

Shared Team Leveling: CD1 and CD8 in the Same Lane

Heroes of the Storm does one thing no other major MOBA does: instead of individual hero levels, the whole team shares a single team level. If I farm experience in my lane, my teammates level up with me. If I go AFK, they suffer for it.

That changes everything. Lane farming, “soaking” in HotS slang, stops being a personal optimization and becomes a sacrificial act. You’re not grinding for yourself. You’re grinding for the team. When my teammate is off in a flashy fight and I’m alone in a lane clicking on minions for boring experience, I’m not missing the fun — I’m doing the work that lets the team have fun.

That’s Core Drive 1, Epic Meaning & Calling, embedded in a supposedly mundane activity. Lane-soaking is the closest MOBAs have ever come to making farming feel like a calling instead of a chore.

But there’s a second Core Drive hiding in the same mechanic. If I don’t soak, the team falls behind in levels. Our enemies hit the level 10 Heroic threshold before we do. They unlock ultimate abilities while we’re still fighting with basics. The whole team is now mechanically weaker, and it’s my fault for abandoning the lane. That’s Core Drive 8, Loss & Avoidance, wrapped around the same action. Do it or we all lose.

CD1 and CD8 braided into one mechanic is one of the strongest behavioral engines there is. The action feels noble AND mandatory. You’re not grinding because you want to; you’re grinding because the team needs you to. You’re not avoiding failure; you’re answering a call. The emotional tone is closer to duty than optimization.

This is why HotS retained a small but ferociously loyal player base long after Blizzard stopped actively investing in it. The game wired pro-social feeling directly into the grind. That’s rare. Most live-service games let grinding feel like grinding.

If you run a business where your team has to do unglamorous background work for the whole org to succeed, like customer success, site reliability, or internal ops, HotS’s lane-soaking is a template. Don’t ask people to grind for personal KPIs. Design the metric so the team visibly levels up together when the work gets done, and visibly falls behind when it doesn’t. The CD1/CD8 braid is the difference between “my quota” and “our survival.” One creates resentment. The other creates loyalty. For a broader look at how good games keep veterans from churning, see The Endgame Problem.

The Quarter-Kill Fish: Not Every Target Is Equal

There’s a character in Heroes of the Storm called Murky, a little fish-man who respawns in seconds and is, by explicit design, worth only one-fourth of a normal hero kill when you take him down.

My in-game reaction: “Chasing him is kind of pointless. He’s only worth one-fourth of a life. So it’s not even worth it.”

Pause on that. The character is designed to be less valuable to kill than other characters. The game is telling you, structurally, “don’t bother chasing this one.” And as a player, I obeyed — I didn’t chase the kill because the reward-for-effort math didn’t justify it.

This is a design lever most people miss. By assigning different values to different targets, Blizzard forces strategic decision-making. Chase the fish for a cheap kill, or peel off and go help the high-value fight elsewhere? Spend effort on Murky, or ration that effort for a target whose death actually moves the game? That’s CD3, Empowerment of Creativity, showing up in a tactical shell.

Now flip it to your business. Not every customer is worth the same acquisition cost. Some support tickets deserve an engineer-hour; most don’t. Some LinkedIn DMs deserve a reply; most don’t. When you treat every target as equal, you get quarter-kill behavior: chasing low-value opponents while high-value opportunities walk past.

The most disciplined teams I’ve advised build the “Murky rule” into their stack explicitly. A named tier of prospects gets white-glove attention; a named tier gets auto-response. A named severity of bug gets paged; a named severity gets deferred. The valuable work is that the designer, meaning you, decides ahead of time which targets are worth chasing. The alternative is an entire team chasing fish because the fish are easy to see and easy to catch, and then wondering why the quarterly numbers missed.

For Designers Who Aren’t Building Games

Six observations from one game session. Every one of them maps to a behavior you’ve seen outside games: in commerce, in employee retention, in social media design, in parenting. The Core Drives don’t care whether you’re playing Heroes of the Storm or running a fintech app. The Core Drives care about the structure of the experience.

If I had to compress the whole session into one meta-insight, it would be this: the rational player is a myth that designers use to excuse lazy design. Every moment I’ve described was a moment where I, fully aware of the irrationality, still behaved irrationally. The horse skin was cosmetic and I felt the loss anyway. Judgment was the weaker heroic and I picked it anyway. The duplicate legendary was supposed to be bad luck and I celebrated the shards anyway. If I, someone with a framework specifically built to dismantle these reactions, still feel them, then your users absolutely do.

The correct response isn’t to lecture your users about rationality. The correct response is to design with their predictable irrationality so the product rewards them for the way they actually behave.

That’s the core promise of the Octalysis Framework. You don’t need to fight human nature to get it to do useful work. You need to understand which Core Drive you’re activating, which one you’re suppressing, and which combinations compound versus cancel. Heroes of the Storm is just one beautifully dense teaching example, packed with simultaneous Core Drive interactions you can watch in real time on any Twitch stream.

The next time you catch yourself thinking “users are weird,” try swapping in “users are predictable.” Pull up the Octalysis map. Find the two or three Core Drives their behavior is actually responding to. Then ask what you’d need to change about the experience to make the same behavior a feature of your design instead of a bug in your assumptions. That’s the work. That’s also the craft, and it’s why I can lose a match, feel bad about a horse skin, and walk away with a blog post.

FAQ

What’s Core Drive 4 in the Octalysis Framework?

Core Drive 4 (Ownership & Possession) covers the motivations built around owning, customizing, and extending the self through objects, pets, identities, and collections. It’s the Core Drive behind cosmetic purchases, character customization, collection sets, avatar gear, and the emotional attachment people form to things that have no utility value. The horse skin example in this post is CD4 at its purest — purely visual, purely optional, and still emotionally weighted.

Why do players choose flashy abilities over strong abilities when streaming?

Because streaming activates Core Drive 5 (Social Influence & Relatedness), which re-weights the value of every choice toward “what looks good to the audience” instead of “what performs best.” The trade-off is predictable: CD4 (identity expression) and CD5 (social payoff) rise, CD2 (accomplishment) falls. It’s not a character flaw — it’s the expected behavior of a rational agent whose utility function includes audience reaction.

What’s the currency preference inversion in reward design?

It’s the pattern where advanced players prefer generic currency over specific prizes, because they already know what they want, and the specific prize might not align. A duplicate legendary skin converted to 400 shards is a better outcome than another random legendary, even though the shards are numerically smaller. The lesson for non-game reward systems: ship currency-plus-catalog instead of random specific gifts whenever the recipient has formed preferences.

How does post-match voting help player retention?

It converts a loss from “my effort was worthless” to “my effort was noticed.” By routing recognition through teammates instead of the system, the losing player walks away with social proof of effort. This matters because competitive games produce losses by structural design — half the players lose every match — so reducing churn-on-loss is a survival issue for the game’s population, not a nice-to-have.

Why does shared team leveling in Heroes of the Storm feel different from individual leveling?

Because it braids Core Drive 1 (Epic Meaning) with Core Drive 8 (Loss Avoidance). Lane soaking becomes a sacrificial act for the team, not a personal optimization. The emotional tone shifts from “grinding” to “duty.” When CD1 and CD8 combine in the same mechanic, you get behavior that feels both noble and mandatory — which is why HotS lane-soaking produces a different kind of loyalty than League of Legends farming, even though the mechanical action looks similar.

Related Reading

• Collection Set Design: The Complete CD4 Guide. The full CD4 playbook for products built around owning and collecting.
• Curiosity’s Dark Side: When Mystery Becomes Addiction. A CD7 counterweight to everything the duplicate paradox teaches.
• The Endgame Problem: How Great Games Keep Veterans from Quitting. A macro version of the post-loss voting insight.
• The Grown-Up Lock: When Ownership and Usage Diverge. A companion post on the ownership/usage gap.
• Actionable Gamification. The full Octalysis Framework laid out end-to-end, with 500+ pages of design patterns and case studies.

WOULD YOU LIKE YU-KAI CHOU TO WORK WITH YOUR ORGANIZATION?