Sponsored By

Quantitative design - How to define XP thresholds?

Increasing number of games include progression trees in their core gameplay experience. The question is how to compute the values associated with each threshold. There exist three techniques. Let’s review them and assess their pros and cons.

Pascal Luban, Blogger

December 11, 2018

3 Min Read

Increasing number of games include progression trees in their core gameplay experience. They allow players to unlock attribute improvements, new equipment or new features. This design mechanism comes from RPGs and MMOs. It makes it possible for game designers to engage players with short to long term objectives.  Such progression trees are very frequent in freemium games but they are also increasingly found in premium triple-A titles. Most of the time, they are based on a « resource » that players have to accumulate by playing the game. This resource can be experience points (XPs) that unlocks thresholds.

The common strategy to define those thresholds is to build a progressive curve. Players quickly level up during the early stages of the game but have to « invest » more and more time to reach higher thresholds, thus increasing their engagement in the game.

The question is how to compute the values associated with each threshold.

There exist three techniques. Let’s review them and assess their pros and cons.


Linear curve

To compute it, simply increase the value for the next threshold with a fixed value. In the following example, each threshold value increases by 3.

This method is the simplest one to implement and to control but it is not without strong shortcomings. Progression could either be too slow during early stages of the game or too fast in later ones. Why? Because it does not take into account the player’s overall effectiveness in the game. In other words, it does not respect the key rule of player’s progression: it does not allow to build a fast progression early on and to slow it down as the player levels up.

Let’s review now a technique that is closer to what we are looking for, the exponential curve.


Exponential curve

With this technique, a threshold value X results from the multiplication of the value X-1 by a coefficient Y. In the following example, the value of Y is 1.4.

This technique is interesting because it makes it possible to offer easy to reach early thresholds and to make later ones much more demanding.

The issue with exponential curves is that it is very difficult to find the right coefficient. If it is too low, the curve will offer a flat progression only. And if Y is too high, values for later thresholds will skyrocket, making them nearly impossible to reach.

There exist a third technique that offers an interesting compromise.


Linear progression curve

Here, it is the increase between each threshold that follows a linear curve, not the threshold values themselves. In the following example, the value increase for each threshold goes up by 1 for each new threshold.

If we compare the three curves, we can immediately see the benefit of the linear progression curve: Early thresholds are easy to reach, then require an increasing amount of points but without reaching hyperbolic values.

To conclude, all curves have their own merits. It is up to you to select the one that fits your progression needs best but keep in mind two rules:

First, don’t hesitate to correct  a computed curve « by hand » , to change values that appear either too low or too high according to your own judgment. This is common practice. Think of Supercell games.

Second, don’t forget that time spent by the player between two thresholds is as important as the values associated with each of them. If it takes too much time for players to reach early thresholds, you are going to loose them. Thus, you must estimate the progression speed of players, how fast they will accumulate the « resource » they need to level up.


My previous blogs:

Does Telltales Games difficulties sign off the death of episodic content business model?

Games that make us think

Far Cry 5: Thumbs up, but whats next?

Subscription 2.0 - Will it become tomorrow's business model?



Read more about:

Featured Blogs

About the Author(s)

Pascal Luban


Pascal Luban is a freelance creative director and game designer based in France. He has been working in the game industry as a game or level designer since 1995 and has been commissioned by major studios and publishers including Activision, SCEE, Ubisoft and DICE. In particular, he was Lead Level Designer on the 'versus' multiplayer versions of both Splinter Cell: Pandora Tomorrow and Chaos Theory, he designed CTF-Tornado, a UT3 mod multiplayer map built to showcase the applications of physics to gameplay, he was creative Director on Wanted – Weapons of Fate and lead game designer on Fighters Uncaged, the first combat game for Kinect. His first game for mobile platforms, The One Hope, was published in 2007 by the Irish publishers Gmedia and has received the Best In Gaming award at the 2009 Digital Media Awards of Dublin. Leveraging his design experience on console and PC titles, Pascal is also working on social and Free-to-Play games. He contributed to the game design of Kartoon, a Facebook game currently under development at Kadank, he did a design mission on Treasure Madness, zSlide's successful Free-to-Play game and completed several design missions for French and American clients. Pascal is content director for the video game program at CIFACOM, a French school focusing on the new media industry.

Daily news, dev blogs, and stories from Game Developer straight to your inbox

You May Also Like