On the basis of the analyzed modern literature, it is possible to draw a conclusion about the novelty of the current study. In the process of familiarization with the literature, a variety of approaches has been identified, which are studied and tested in the practical part. According to them, the question of the research is: Which aspects should be considered while investing in blockchain projects? The following parts provide a detailed answer to this question, as well as all the necessary comments and developments received in the process. Over the centuries-old history of trading, investors have formed a huge number of investment strategies that are used in certain moments of trading. They are actively used in the traditional field, but can they be used when investing in blockchain projects? This division should be introduced because of a completely different specificity of trading on cryptocurrency exchanges. Fast and sharp fluctuations in the prices of crypto assets force investors to reconsider their behavior. Even well-established strategies in the "crypto" market can be ineffective or require updating over time. The primary difference of this work is a high degree of relevance because the data were obtained during practical work in a crypto company. The stated objective of the study also corresponds to the question, revealing it more specifically from the practical side. The objective is to study what are the main factors considering investments in blockchain startups and to make a guide that will help to avoid losses while manager works with blockchain. These factors include investment strategies, as well as all sorts of markers and metrics by which a person can make an assessment of the feasibility of investing in a particular project. As a rule, open information may be used to make a preliminary conclusion about the project. These may include capitalization, the project team, the essence of the project, deadlines and so on. In this study, the selection and evaluation of these factors are carried out in order to determine the most influential for subsequent investments. On the basis of such factors, we can create a guide- a cumulative instruction, according to which it is possible to estimate the investment attractiveness of a particular blockchain project. The use of this guide should reduce the likelihood of investment in a failed project, and therefore save the investor's funds. Of course, the use of this guide is advisory in nature and is unlikely to replace the complex system of evaluation of the largest investors, but its potential audience refers to private investors without huge funds for investment. This form of the practical guide is designed to help form a basic understanding of the key parameters for the evaluation of blockchain projects. Its use does not require special technical skills and prior scientific work. Understandable for perception language and logical scheme of drawing up are aimed at facilitating the understanding of the meaning of what is written. This guide allows answering the question of the study, as it contains information about what to do and what criteria to pay attention when investing in blockchain. Certain hypotheses were formulated for further research, including in-depth interviews and statistical analysis. Their verification is carried out on the basis of these research methods. These in-depth interviews are conducted with cryptocurrency investors and crypto enthusiasts who have different trading behavior. All of them have a unique experience in a particular direction. Our questions were developed for every participant individually. The procedure of the interview was formalized and will be described in further parts. Some respondents were comfortable to talk only in English, the other part only in Russian. Accordingly, the language of the interview varied depending on the capabilities of the respondent. We recorded the interviews in the native language of the respondents. To match the language of the work, the Russian texts have been translated. Emphasis was placed on the general meaning of the above, as well as key answers that confirmed or disproved specific hypotheses. According to the answers of the interviewed, as well as statistical analysis, we can check the following hypotheses, and decide whether to accept or reject them. The hypotheses are based on the studied literature and founded a gap in previous studies. The team members worked directly with the ICO, namely with their assessment, selection, analytics, investments. This experience helped to formulate the hypotheses. Many of us are involved in the crypto sphere financially; we also have a certain set of cryptocurrencies in the "crypto" wallet and regularly follow the rates and news. This information is provided to facilitate the reader`s understanding and help to formulate a certain opinion about the research team and its competence. Returning to hypotheses, this study checks the following hypotheses:

- Short-term strategies are more effective in crypto the industry rather than long-term ones.

One of the popular opinions is that cryptocurrencies are profitable in the short term. For example, immediately after the start of trading a new currency, many investors are trying to sell it immediately above the purchase price, without holding it for a long time. Despite this, there are strategies for long-term storage of certain currencies. It is assumed that these coins in the future can rapidly add to the price, which will gain profit that overlaps long wait and other costs. The study compares data on both sets of strategies to identify more effective ones.

- The presence of "High public bonus" is better for the profitability of the blockchain project.

Companies try to draw investors` attention using a lot of instruments. Usually, the cost of one conventional unit purchased on the very initial phase is low. This is due to increased risks for a number of reasons. First, there are a lot of scammers who can mislead people by inventing non-existent projects. After collecting funds, such projects are closed, and investors are left with nothing. In addition to direct scams, there are a number of reasons that can affect the project on the very initial phase. However, participants receive an increased benefit. One of them is a public bonus. This is a rate of additional tokens. Investors can receive it at the very beginning of tokens` sale. It means that an investor can receive more tokens for the same price. The bigger value of public bonus the more additional tokens a company gives. Respectively, it is logical to participate in projects which give extra tokens for free. This was not checked by previous works; our research team decided to include it into the analysis.

- Investments in projects which were already chosen by funds are profitable for ordinary investors.

Since the appearance of cryptocurrencies and crypto-related projects, their number has increased significantly every year. The peak came in 2017 when most cryptocurrencies were growing in price rapidly and inexplicably. At some point, the growth stopped and began an equally rapid decline. This was one of the factors formed the opinion that blockchain projects can make a profit if someone buys coins at an early stage, especially funds which can afford a large sum of money to invest. It is needed to closely monitor the rate and quickly sell at a favorable price. Ordinary investors remain uncertain about the long-term prospects of blockchain projects. At the same time looking at the funds` best choices may significantly simplify analyzing the process for amateur investors at least at the initial stages when there is no much experience in the industry.

- "Blockchains" and "Blockchain service" categories of ICOs (infrastructure ones) are more preferable for investments related to other categories.

Information noise around the word "blockchain" can play a significant role in the success of the ICO. There is an opinion that it is profitable to invest only in "blockchains" because they are considered more stable and efficient comparing to other categories. This logic looks quite adequate because the popularity of the project has a positive impact on the success of the project. The more people know about it, the more likely it is to attract investment and successfully implement it. To raise awareness and draw attention just might become a popular use of blockchain technology. In addition to blockchain categories of projects choosing to held ICO there are also other projects with different category also based on blockchain which are also given attention to ensure an integrated approach to the study of the issue.

- The degree of hype around a blockchain project positively affects its profitability for investors.

Major changes have occurred in the cryptocurrency market. We have seen various market cycles from general price and capitalization growth to a rapid decline and correction. All these periods relate to the mass popularity of blockchain technology and even its prestige or trend. The more rush around the project, the more people can hear about it. High-profile projects often attract the attention of investors. Accordingly, if the project is known and popular, it is more likely to be successful at the fundraising stage. It also reduces the likelihood of fraud from the company. The popularity of the project can affect its stability. But how "hype" indicator may affect the revenue of the investor? Verification of this hypothesis answers this question. This is possible because of the presence of "hype" metric in our database. This hypothesis implies that it is more profitable to invest in blockchain projects with high hype level.

- There are certain indicators of ICO which negatively influence on projects` profitability.

Every ICO team publish a lot of information about their project. This data is presented in open sources and special websites. Before this moment, we thought about the indicators and metrics that could help us to see potential earnings of investments. We assumed that there are key metrics that influence projects in a good way. Same time, this hypothesis implies that there are criteria that could mark less successful or even unprofitable projects. It also suggests that we can make an investment decision based on the definition of negative markers of the project. Accordingly, if such markers are present, is it worth investing in a particular project? This hypothesis is tested on the basis of statistical analysis. Using statistical methods, we will check the relationship between the main ICO criteria and its profitability.

To test these hypotheses, as well as the study of the research question, quantitative and qualitative methods are used. This distribution is due to the specifics of the selected topic. Most of the previous studies were qualitative and contained expert assessments and other data of a qualitative nature. In this study, the methods of qualitative research are also used, but this does not deprive it of its value. This approach is important in order to collect expert assessments at the current stage. The crypto-related sphere moves very quickly, the conditions of the game can change several times during the year, which means that past quality data can quickly lose its relevance. In order to track what investors think at the moment, as well as to obtain their expert forecasts for the future, using in-depth interviews in our study. Interviews were conducted offline, participants are anonymous, the results are presented in the practical part. In addition, quantitative methods are used in the work. This is quite a rare case for the crypto-related sphere because complete databases are not presented in the public domain, which greatly complicates the data collection process, and therefore the subsequent analysis. This research paper includes several statistical analyses as quantitative methods. The basis of the statistical analysis is the database of all ICO for 2018. Often, obtaining this data is a big problem for researchers, because of their lack of open access. As a rule, companies working in the crypto-related sphere and leading certain statistics do not go to give hard-earned achievements to strangers. This behavior is understandable, since such materials are of commercial value and it can significantly improve the position of competitors. It was significantly useful that one of our researchers participated in the compilation of such a database earlier. The research team got a summary table of all ICO, which took place in 2018. In addition to the names and terms, it includes a huge number of parameters that represent a great potential for research and analysis. A large summary table presents both descriptive parameters and financial indicators. It can be types, country of origin, the number of members of the development team, platform, and so on. Financial indicators such as total funds collected, percentage of profitability, and others are also included in the above table. The "R" program was chosen as a tool for statistical analysis due to the relative simplicity of the work, as well as the rich functionality of the program. It is one of the most popular and widely recognized among scientists' tools for data storage and Analytics. Details of the process of work in the "R" and research methods are described in the following parts of the research work. The work in this program was reduced to the analysis of individual ICO parameters contained in the general database of all ICOs for 2018. As a result of the analysis and further work with the data, our team constructed a linear regression model describing the set of all analyzed criteria. Some of the above hypotheses were tested using these statistical tools.

Chapter 4. Methodology

Methodology in the research paper aims to supplement the theoretical basis discussed previously and to describe and justify the selected methods. This section provides a detailed description of the research methods, a description of the hypotheses, their relationship with the theoretical part. We propose several hypotheses regarding our topic to accept them or to decline. The hypotheses were formulated based on the analyzed literature. At the moment, most studies on blockchain projects are informational in nature. They test the blockchain technology itself. As a result of the examination of the theoretical materials and scientific works, the study of certain aspects of ICO was not fully revealed. At the same time, many authors note the importance of certain aspects. For example, they write about the importance of the blockchain project` team. Meanwhile, mathematically this importance is not described. Besides, our interviews with experts in the blockchain-related industry which were aimed to supplement all the obtained information have also shown that despite all the importance they incur, such form of data gathering still have a qualitative nature and needed to be checked additionally to be statistically proven. In order to check whether there are mathematical relationships between these aspects and the investor's revenue, certain hypotheses were also made. Our stated hypotheses are:

a) Short-term strategies are more effective for ICO investments rather than long-term ones;

b) Investments in projects which were already chosen by funds are profitable for ordinary investors;

c) "Blockchains" and "Blockchain service" categories of ICOs (infrastructure ones) are more preferable for investments related to other categories;

d) The presence of "High public bonus" is better for the profitability of blockchain project;

e) The degree of hype around a blockchain project positively affects its profitability for investors;

f) There are certain indicators of ICO which negatively influence on projects` profitability.

Research is consists of several parts. Firstly, the work is related to the conceptual component of the new technology, namely "blockchain". According to Warren Buffet for investors, it is crucial to understand what they invest in. Then there was active testing of the stated hypotheses. For instance, the research team has attended major blockchain and cryptocurrency events such as Blockchain Life and Ian Balina meet-up. During the meet-up, two interviews were held. This helped the research team to form more precise picture regarding investors` opportunities and possible risks. Several research tools are used in the work: three in-depth interviews with representatives of blockchain related category of businesses. It was vital for the research team to understand what methods of project evaluation respondents use, what tactics they have on the return of investment (to keep or, for example, immediately withdraw). The block of questions has included the main characteristics of project evaluation which can exclude unpromising projects in terms of ROI (with the highest risk score) already in the initial stages of the analysis. After conducting the interview and identifying the overall results, we analyzed all the gathered data.

Speaking of methods, both quantitative and qualitative methods are used. The qualitative method is represented by an in-depth interview. This method helps to receive actual information from experts in this blockchain related industry. A characteristic feature is the small number of respondents. In-depth interview is an appropriate method to get information about the personal experience of the investors. It also provides comprehensive data to analyze because of the unlimited speech of a respondent. Unlimited means that there are no close questions, the respondent formulate the answer in a personal manner and then it should be interpreted carefully. The interpretation depends on the researcher`s mindset, so there is a risk of distortion. Our research team has worked on accurate formatting the questions to avoid such type of risk. This method has a lot of advantages. The main advantage is that in-depth interview provides more detailed information than other data collection methods. However, at the same time, it has a set of limitations. For example, the quality of the gathered data depends on various factors. These factors may be the mood of the respondent, his or her understanding of a question, skills of an interviewer, conditions of the interview`s process. Also, the sample of the interview is small, so it is impossible to generalize the results. Despite the limitations, the in-depth interview remains quite efficient and suitable method concerning our research. Cryptocurrency investors are valuable respondents because of the unique personal experience in this particular sphere. There are different stages of conducting in-depth interviews. Firstly, we developed a plan. The plan is to identify the investors to ask; to specify the topic and set the needed information. Further steps are connected with questions and reviewing process. It includes instructions for the interviewer. For example, what to say at the beginning of the interview and how to record answers; in general, how to behave and lead the interview. One of the main preparation steps is the development of questions for the interview. They were not the same for all respondents taking into account the specialization of the investor. It means that every participant has a personal set of competences. According that, the list of questions should fit these competences to better develop the necessary field. As a result of the in-depth interview, we receive records of the interview. Further, the research team work on raw data and transform it into a completed note of the interview. These notes are added in appendix 1.

Moreover, we have made a statistical analysis of blockchain projects over 2018. The main emphasis is placed on the characteristics of the blockchain projects` selection. Quantitative research methods were used primarily by statistical data analysis of different blockchain projects.

Statistical analysis consists of testing hypotheses using statistical methods, as well as building a linear regression model. To consider these methods in more details the original database of ICO projects was used. Metrics are collected for each project. The analysis of these metrics allows to check the stated hypotheses. This test is based on the analysis of the relationship between the selected independent variables and the dependent variable "Return test". This indicator means the percentage of profitability of the project after the opening of trading. Accordingly, it is calculated using the ratio of the average percentage increase in tokens` price for each of the first 5 trading days to the price of the token before the opening of trading. To check the dependencies between the dependent variable "Return test" and dichotomous variables (categorical variables which take only two values) we used the Mann-Whitney test. (Ruxton, 2006) For categorical variables with lots of values, a pairwise Kruskal-Wallis test. First, consider the Mann-Whitney test. This is an analogue of the t-test, its use is justified by the fact that our dependent variable is not normally distributed. There is a categorical variable, it takes only two values - yes and no. This test compares the average values of the "Return test" with a quantitative variable. As a result, this test shows whether the difference in the average values of the "Return test" between the two categories "yes" and "no" are statistically significant. Accordingly, only statistically significant data are presented in the results. For variables with more than two categories were used Kruskal-Wallis test. This test is similar to the previous one and compares each value with each other in pairs. If the value differs significantly, it is displayed. For the remaining data, Spearman rank correlation coefficient was considered. It was used to calculate correlations between "Return test" and quantitative variables. This algorithm assigns orders starting from 1 and more to the data. Similarly, the order is assigned to the dependent variable. Then, the algorithm looks for a correlation between these orders. Based on the analysis, we build a linear regression model. Its structure is demonstrated in table 1. The metric field specifies the explanatory variables that affect the dependent variable. The Estimate field indicates the degree of influence of the independent variable on the dependent variable. Accordingly, a positive value indicates a positive correlation, negative values a negative correlation. The algorithm calculates the Estimate value, which is the coefficients , in the equation

.

Y is the dependent variable, x is the independent variables, is the intersect. Other parameters correspond to the section name. Pr is the probability that the null hypothesis is true, and that the estimate is not significantly different from zero. If the p-value is less than 0.05, the estimate is significantly different from zero. Depending on the sample size, this threshold of 5 hundredths can be extended. If the sample is small, this threshold can be increased.

Table 1

Structure of the regression model

The statistical analyses were conducted on the basis of the ICO database. This database contains different metrics for all ICO 2018. The format of this database is a pivot table. For clarity, there is a transcript of the parameters that it describes. The initial code of "R" is in appendix 2. The structure of the database description is based on columns and their alphabetic headings. The description is from the beginning of the database on the left to the end metrics on the right. In general, the data is a continuous table. For clarity and to preserve the logic of the narrative, the columns below have been broken down to take into account the available space of the A4 page. In the original table over 2000 values were presented. This section lists the first 10 lines with examples of actual project data. As a part of the guide, our research group has developed an Excel table, which helps to represent all the needed data. All the parameters that, in our opinion, are worth mentioning are included.

Description of the data.