At 8 pm on January 19th, Voodoo conducted the fourth Chinese live broadcast at station B (search Channel_Voodoo) and brought the theme of “Knock’em All Case Sharing + How to Record a Good Test Video”.
The following is a compilation of live content, I hope it will be helpful to all developers.
1. The OHM Studio for All People
“Knock’em All” was brought by OHM Studio, which is a French studio established in 2018. Up to now, OHM and Voodoo have successfully released 5 products, namely “Shortcut RUN”, “Knock’em All”, “Push’em All”, “Bullet Rush!” and “Knock Balls”. According to third-party platform statistics, the cumulative downloads of all OHM products have exceeded 280 million times.
Before “Knock’em All” there were two very successful products: “Knock Balls” and “Push’em All”. So OHM combined the names of the two products to make “Knock’em All”. In fact, it can be seen from the name that “Knock’em All” combines the gameplay of the two products, and extends the pleasure of catharsis in the two games and the pleasure they call “guilt” into the game.
Why share this case? Usually in Voodoo’s message to developers, the first one is not to try shooting games, because this is a masculine theme, which may not meet the needs of the mass market. But “Knock’em All” “slapped the face”.
“Knock’em All” brings shooting, which is rarely seen in the ultra-casual game, to the game, explores new areas, and tries the first-person perspective. These elements have not been seen before “Knock’em All” of. This is a bold and innovative case, and it also raises a question: Can so-called masculine games not be made into hyper-casual games?
We see that common shooting games emphasize aiming and shooting, but in “Knock’em All”, it is more physical-the physical feeling of a bullet falling, the physical feeling of the puppet system. After these physical senses are merged together, it is discovered that even in the same level, because of the physical sense, each time you play, you have a different experience, which increases the replayability of the game. This kind of creativity also proved that the depth of the game can continue to deepen.
So we encourage everyone to think about the depth of the game during the creative stage. “Knock’em All” is a very good case.
And I want to tell everyone that OHM has a full-staff creative conference every week. Every employee must put forward ideas, including art, development engineers, students, etc. must put forward ideas for discussion, and make selections based on a large number of ideas. I believe this is also theirs One of the ways to succeed continuously. Moreover, art and development students have a unique advantage in their creativity, that is, it is easy to jump out of the vicious circle of designing games for “game players”. Their creativity may be more in line with the thinking of non-game players and more “super casual”!
2. “Knock’em All” tuning and review
After determining the idea, OHM carried out rapid prototype development and conducted the first version test. There are 10-20 manual levels in the V1 version. Manual level design does have a certain amount of work, but as mentioned earlier, “Knock’em All” is a fusion of the two products, so the past materials can also be reused to improve their efficiency.
The early version has limited content and only made one enemy and one weapon. The reason is simple. The first version aims to test whether this gameplay is accepted by ultra-casual users. It can be seen that even if there is only one enemy, one weapon, and 10-20 levels of manual levels, it is enough to cope with the first version of the test. And the very linear and intuitive conversion of the first version does not have more level design. The result of this test is a CPI of 20 cents, 46% of the time stays, 15% of the seven stays, and 20 minutes of game time.
What I want to emphasize here is that the first version test hopes that everyone can carry out rapid prototype development, because as long as the gameplay can be verified, then the purpose of the first test is achieved. If the development of the first version exceeds two weeks, then you need to reflect on whether your creativity is in line with ultra-leisure, and whether there is a problem with development efficiency.
In the second version, because the data proves that the gameplay is accepted by users, more attention needs to be paid to content and diversity. This version has invested time in designing more manual levels, adding more enemies and weapons, including the introduction of Boss levels. These contents have effectively helped the second and seven stays of the game.
We suggest that it makes sense to make these efforts only after the gameplay is verified. In addition, “Knock’em All” has excellent level loss data. Although the long-term retention data of the game is general, when doing level analysis, it is found that the churn rate between levels is less than 5%. This shows that most users accept the current difficulty and gameplay. Developers only need to provide more content. can.
After the two versions were tested, they were soon confirmed to be online, because the game’s data is good enough. After it went live, the game added more Meta systems and economic systems, including skins and currencies. And created a “dismemberment effect” to give users a stronger sense of refreshment. The “dismemberment effect” affects both CPI and retention. In addition, Boss levels and environmental interaction have been added.
1. Although it is a shooting game, it does not seem to have the difficulties of traditional shooting games. Instead, it retains the fun of shooting games and makes him look less like a shooting game. This is what we think OHM has done very well. At the same time, Boss levels can give users a more refreshing experience, but it is not necessary to beat the Boss. This is also one of the designs to reduce the difficulty and make the game more suitable for ultra-casual users. Not only that, OHM applies the development experience of “Push’em All” directly to the game, and the user’s shooting can interact with the environment, which increases the replayability and diversity.
In the second version on line, game sound effects were added based on user feedback. It has to be said that adding sound effects to shooting games can indeed help data increase, but we did not add it until later. Because there are trade-offs in the early stage, the focus of the early stage should be on the construction of the core loop.
At this stage, OHM continues to explore, adding flame mode and parkour mode. All of this has a great impact on the game duration. iOS game duration reaches 30 minutes, and Android game duration reaches 40 minutes. This is rarely seen in hyper-casual games.
In summary, what we want to share about “Knock’em All” is:
1. How to reduce the CPI of ultra-casual shooting games? We recommend zooming in on shooting fun. “Knock’em All” is not violent, but it makes users feel that they are very powerful, can destroy everything, and can defeat one hundred. We think this may be a viable path for ultra-casual shooting games.
2. Driven by physical effects, the replayability continues to increase. The shooting trajectory and the feedback of the puppet system to the shooting make every game feel fresh. This physical effect is always recommended by Voodoo.
3. The ultra-casual game studio should be creative with all staff, first ensure the quantity of creativity and then improve the quality. This is what we have seen in common between OHM studios and many Voodoo studios that can continuously build explosive game studios.
3. How to make a good test video
The test video is placed on Facebook. It is a relatively short video to attract viewers and let them choose to download your game. Here I call them audiences, not players, because most ultra-casual users are not actually game users, so for them, each game is an advertisement first, and then a game. This is very important and will affect the process of thinking about creativity.
How does it affect? When you come up with a new idea, you can ask yourself: What does this creative advertising video look like? We have seen a lot of great ideas, and these ideas will produce great games. But these creative test videos are boring or too complicated. This is the problem of the core gameplay of creativity. Remember, in the field of ultra-casual games, even if the game is great, if the CPI is too high, it is not a good ultra-casual game.
Because our test platform is self-service, the studio needs to record its own test video.
We recommend that the test video is 10-25 seconds, less than 100MB. We found that viewers only watch the first 4-6 seconds of the ad, which means that the core gameplay of the game must begin immediately. There should be no menus or splash screens in the video, and no editing. No need to make game trailers, just a pure game video.
Be sure to put the coolest feature of the game in the first 4-6 seconds, and an interaction must be displayed within 4 seconds. What is interaction? It can be how the player interacts with the game, it can be how the character interacts with obstacles and enemies. In addition, everything in the video must be clear. The audience needs to be able to see who I am, what I can interact with, what is the obstacle, what is the path, and what is the background.
The following example is “Fat Pusher”. The CPI in the first test was not ideal. I think the reason is that there are many useless things in the background. Although these things look good, we think they distract the audience from the core gameplay. So in the second test video, the developer focused on reducing the visual disturbance in the background. After the change, the CPI was significantly reduced, and it was close to the issuance standard.
In the third iteration, the video showed an interaction earlier and used a more realistic character image. In addition, this version changes the lens angle. Finally received a CPI of 0.15$.
So how to record game video? In fact, there are many recording tools on the market. Vertical or horizontal screen is fine, we are willing to test any kind of game, any kind of video. But remember, all the ultra-casual games on the recent leaderboards are in vertical screen, so we all recommend vertical screen. Finally, make sure that there is no black border around the video, and all the videos are game screens.
In one sentence, it takes 4 seconds to show how the game is played. In detail, it is the coolest point in the game. The mechanics in the game must appear in the first 4 seconds. As mentioned earlier, the developer of “Stair Run” cut the first 3 seconds of the test video during the second test, because he believed that the most important thing is to build a ladder, so he started to show the ladder from the first second of the video. The mechanism immediately reduced the CPI by 50%.
This is what we emphasize, we must start to show the most interesting part of the game in the first second, there is no need to show the game menu, splash screen, or even editing, because this is not a game trailer.
Show the interaction within 4 seconds, so that the audience can participate in the game and understand the game. Therefore, the premise of a low CPI is that the user understands the game in the first 4 seconds, and then can play it after understanding. So grab the audience’s attention in the first few seconds. The most interesting part of the game is what can grab the attention.
1. Direct recording
2. Show the core fun in the first second, let users understand in 3 seconds
3. No need to show the game menu
4. It is unnecessary to show the game’s victory and defeat (optional)
5. It is recommended that you can prepare 3 videos at the same time to test together. You can choose different lens angles, different colors, and different challenges. In this way, we can obtain more information during the CPI test, and improve the accuracy and efficiency of our decision-making.
Four, frequently asked questions
1. When should iterate?
Our issuance standard is that the CPI is less than 0.2$, the second stay is higher than 45%, and the seventh stay is higher than 12%. For this reason, we also have a 30/30 principle-if the first test result is a CPI lower than 30 cents, and the second stay is higher than 30%, we think the product is worth iterating at this time. Otherwise, we suggest pivot or cut off the project and devote energy to new creative research and development.
2. What is Pivot?
Pivot means “axis”. If we see data that is not good enough during the test, but there are shining points-such as CPI 20 cents, but the second stay is very bad-we should analyze the reasons for the flashes in the product (what makes the CPI so low), and the existence Question (What makes the retention difference?) Then, change the angle (flip along the “axis”) to improve it. This is what we call pivot, or big change. The premise of the big change is to keep some things and change some things. This is our understanding of pivot.
Supplementary note: Pivot is usually used when the CPI is low, but the retention is low at the same time, because the CPI is difficult to reduce, but the retention is relatively easy to increase.
3. CTR or CPI test?
We recommend the CPI test because the CTR test provides a low probability of CPI. If a 5% CTR is measured, it means that the chance that the CPI is less than 20 cents is about 45%. The decision to be made at this time is to continue or give up. If the CTR is less than 5%, choose not to do it at this time, in fact giving up a 20-40% CPI below 20 cents. It still looks very risky, doesn’t it?
CTR provides probabilities, not exact values. Making decisions based on this probability is a risk and there is greater uncertainty.
In contrast, the CPI test provides more information to help us make decisions. First of all, he provides direct, accurate, no conversion, not probability, but accurate data CPI. And the CPI test can also obtain the second and seventh retention data, which is convenient for making better judgments. There is also the length of the game, whether the user likes it, all can be obtained from the data. And we recommend that developers track the progress of the level, and the loss of the level determines whether the user accepts it.
Some developers say it’s okay to do the CTR test first and then do the CPI test? Yes, but you will find that such a process will waste 8 working days, because it takes about 8 working days from video recording, to testing, to getting the results, and then 30% of the time will be wasted in the development version, in other words , Lost 30% of the winning rate. Moreover, you will find that if you can shorten the time of prototype development, it will be more attractive to conduct CPI testing directly. So, here is a question for everyone to discuss: Is one of the reasons for the CTR test because the development efficiency is not high enough?
Hope it is useful to everyone, the above is this sharing.