Why Beginner Coders Building Online Games Should Expect Small Wins First
A new member joined our Thursday club last spring with a clear plan. In six weeks he would build a multiplayer online game with a shop, ranked matches and a reward system "like the big ones." By week three he had a window that opened, a...

A new member joined our Thursday club last spring with a clear plan. In six weeks he would build a multiplayer online game with a shop, ranked matches and a reward system "like the big ones." By week three he had a window that opened, a square that moved, and a growing sense that he was failing. He was not failing. He was exactly on schedule for a beginner, and nobody had told him what the schedule looked like.
Players who come to coding from games often carry a picture of outcomes shaped by what they see online: the clip of a giant win, the streamer who levelled up overnight, the page promising a huge result for little effort. One of our mentors keeps a short list of such pitches she shows new members, ranging from "learn to code in a weekend" courses to game boosting ads and pages like ankertoto, and asks them to spot what all of them leave out. The answer is always the same: the distribution. They show the rare big result and hide the hundreds of ordinary ones around it.
This article is about resetting that picture, in two ways. First, for your own progress as a coder. Second, inside the games you build, where modelling a "big win" honestly turns out to be one of the best beginner exercises we know.
What Progress Actually Looks Like in the First Months
We tracked the projects of 27 club members over their first twelve weeks of Python. Their paths varied, but the milestones landed in a fairly consistent order:
- Weeks 1 to 2: printing, variables, simple input. A text guessing game is a typical first win.
- Weeks 3 to 5: loops and conditions. A dice game or rock paper scissors with a score counter.
- Weeks 6 to 8: functions and lists. A small text adventure with an inventory.
- Weeks 9 to 12: a graphics library such as Pygame. One moving player, one obstacle, a game over screen.
Not a single member had a working online multiplayer game by week twelve, and that includes several who were very quick learners. Networking, synchronising state between players and handling disconnects are each bigger topics than everything in the list above combined.
The member from our opening story finished his first term with a single player arena game: one character, three enemy types, a score saved to a file. It was a genuinely good result. Once he saw it as one, his motivation came back.
Why Small Wins Compound
Small projects are not a consolation prize. They are how skill accumulates. Every finished project teaches something a half-finished big one does not: how to end a program cleanly, how to fix the last annoying bug, how to explain your code to someone else. Our piece on how making a simple game teaches you to code goes into why finishing matters so much.
There is also a practical reason. A beginner who attempts a huge project usually stalls at the first hard problem and has nothing to show. A beginner who finishes five small projects has five things to show, and the fifth one is usually far better than the first.
Modelling a Big Win Honestly in Code
Here is the exercise we run in week five. Members build a simple reward chest of the kind found in many online games. The chest gives one of four outcomes:
| Outcome | Coins | Chance |
|---|---|---|
| Common | 10 | 70 percent |
| Uncommon | 40 | 22 percent |
| Rare | 200 | 7 percent |
| Jackpot | 2,000 | 1 percent |
Step one is to write a function that picks an outcome. The simplest version draws a random number between 0 and 1 with random.random() and walks through the cumulative chances: below 0.70 is common, below 0.92 is uncommon, below 0.99 is rare, and anything else is the jackpot.
Step two is to open the chest 100,000 times in a loop and record every result. Step three is to calculate the average coins per chest. Members are always surprised by what comes out. The theoretical average is 7 plus 8.8 plus 14 plus 20, which is 49.8 coins. The simulated figure lands within a coin or so of that.
Then we ask the real question. If a player opens 30 chests, how often do they see the jackpot at all? Members write a second loop that simulates 10,000 players each opening 30 chests. Roughly 74 percent of those simulated players never see a single jackpot. Most of them end up with far less than the headline number suggests.
What the Simulation Teaches
This short program covers loops, functions, lists, counting and averages. It also teaches something no syntax lesson can: the difference between the best possible outcome and the typical one. The jackpot is real. It is also rare enough that most players never touch it in a normal session.
Several members have gone on to build histograms of their results with a plotting library, which turns a column of numbers into a picture of where most outcomes actually fall. It is a striking image: a tall bar for small rewards and a tiny sliver far to the right. We connect this to the club's earlier work on how random number generators work in games, so members understand that the computer is not "deciding" to be generous or stingy.
The Advice We Push Back On
A lot of beginner content says "think big, start your dream project now." We understand the appeal, and ambition is not a bad thing. But we have watched that advice cost people months. Starting a dream project in week two tends to produce a folder of abandoned files and the belief that coding is not for you.
Our advice is the opposite: keep the dream project in a notebook, and break it into pieces you can build alone. A multiplayer shooter becomes a moving player, then shooting, then enemies, then a score, then saving, then, much later, two players on one keyboard. Each piece is a small, finished win. Networking can wait until the rest is solid.
Setting Expectations as a Mentor
If you run a club, say the honest timeline out loud in the first session. Show the milestone list above. Show a finished week twelve project from a previous member, so newcomers can see something achievable rather than a commercial game. Celebrate finished projects publicly, however small.
It also helps to normalise frustration. Every member hits a week where nothing works. Telling them in advance that this week will come, and that it is part of the normal curve, keeps more of them in the room. We wrote more about this in getting unstuck on a practice problem.
Small Wins, Honestly Counted
The member who wanted to build a multiplayer game in six weeks is now in his second year. He has built a networked two player game, with a lot of help and a lot of rewritten code. He says the reward chest simulation was the session that changed how he thought, both about code and about the games he plays. Big wins happen. They are just rarer, and slower, than the clips suggest.
More in Games
Games
How Coding a Number Draw Game Teaches Probability
Probability is one of those topics that stays slippery on paper and suddenly makes sense the moment you code it.
Games
How Coding a Game Trivia Quiz Teaches Python Dictionaries
Dictionaries are where a lot of beginners stall. Lists make intuitive sense because they are just things in a row, but a...
Games
Why Coding Game Hangman Teaches Strings and Loops
Ask a new programmer to practice string handling and their eyes glaze over.