COMBINING DIFFERENT MCDM METHODS WITH THE COPELAND METHOD: AN INVESTIGATION ON MOTORCYCLE SELECTION

There are many different multi-criteria decision making methods in the literature. These methods, which enable criteria with different measurement units to be examined together, allow choosing between alternatives. However, different methods can produce different results depending on the data set. The aim o f this study is to combine the results obtained by applying different methods to the data set with the Copeland method. To this end, a problem with real data was first addressed. Technical data o f motorcycle alternatives that can be preferred for individual needs were collected in terms of different criteria. The weights of these criteria were found by the PIPRECIA method. Six different multi-criteria decision making methods were used to evaluate motorcycle alternatives. These methods are MOPA, MOOSRA, COPRAS, SAW, WPM and ROV. The sequencing results obtained from these methods were combined with the Copeland method and the results were discussed.


Introduction
Technological developments and innovations have changed the lifestyle of people today, which in turn increased the competition between firms that produce for very big markets. Therefore, even tiny differences between end products which are better suited for customers' needs can create a competitive advantage for the competing firms. Motorcycle market is a market that reaches high selling numbers and high financial volume (Yontar and Aras, 2018). Especially, sport motorcycle take attention due to their higher prices and higher technologies (Bilgin, 2018). Sport motorcycle market is considered as a niche market in which many firms compete and create new and improved products. Sometimes as a hobby and sometimes as an adventurous desire, sport motorcycle fans follow the developments in the market very closely. They do look at some of the criteria when comparing different models. These can include technical specifications of the models or more subjective elements such as design.
In this highly competitive market, firms offer such products that are very close to each other in terms of criteria customer desire and in terms of performance. Different motorcycles offered by competing firms with close features can create confusion for sport motorcycle buyers.
Suzuki, Honda, Kawasaki and Yamaha are considered important firms in this market. BMW is another established producer, however in this study, due to closeness between the origin of the firms and offering similar features, first four producers are considered. The models are chosen among alternatives with motor capacity under 1 litre. Technical specifications of the selected models are taken from the official Turkish website of their respective firms while the prices considered are dated to the time the study was conducted.
The criteria which affects buyer decisions are found out by conducting interviews with sports motorcycle fans. The weights of the determined criteria are calculated with the PIPRECIA method. The four different firms which sell to the Turkish market are evaluated individually with multi criteria decision making methods of MOPA, MOOSRA, COPRAS, SAW, WPM and ROV. The results of these different methods are combined with the Copeland method and the outcome is discussed.

Literature Review
There are studies conducted in different topics of motorcycles in literature. Some examples from these studies include: Risdiyanto et al. (2020), investigated people's choices for transportation between motorcycles and moto-taxis in Indonesia. Akarpa and Diler (2020) has taken an artistic approach to motorcycle design. In this context, they studied the usability of style and art fashions in design process of "CBF 500" model. Rojniruttikul

Methodology
In this chapter, the procedures of the methods in the study are explained with the equations. Table 2

( criterion j is important than criterion (j -1) ^ Sj > 1 importance o f j = importance o f (j -1) ^ Sj = 1 criterion (j -1) is important than criterion j ^ Sj < 1
(1) Table 3 shows the process of MOPA method.  The last step in Table 3 shows the values of the alternatives in the problem. The highest value shows the best alternative. Table 4 shows the process of MOOSRA method.

Finding of the midpoints Ui = u++и-2 (35)
Source: Madić et al. (2016) The highest midpoint shows the best alternative according to ROV method. Table 9 shows the procedure of Copeland method.

m: the total number o f MCDM method k: rank value according to MCDM method n: total number o f alternativestoplam alternatif sayisi i: rank value o f alternative in the row j: rank value o f alternative in the column rk(Af): rank value o f alternative i according to method k
Ozdagoglu et al.

fk (i,j): superiority o f alternative i over alternative j S (i,j): overall superiority o f alternative i to alternative j G(i,j): win, loss and tie conditions o f alternative i over alternative j
GPp win score alternative i

УРр loss score fo r alternative i
СРр Copeland score o f alternative i

Win, loss and tie conditions r S (i,j) > S(j, i) Ai Ф j ^ 1 G(i,j) = { S J J^ = S J , j A i^j^1
(38)

Application
In order to integrate the results obtained from different MCDM methods, firstly, a problem in which the real values can be found was tried to be determined. For this purpose, the problem of choosing a motorcycle for motorcycle enthusiasts is discussed. Within the scope of the study, four sports motorcycle alternatives belonging to Far East brands sold in Turkey were evaluated: The alternatives are as follows.

Alternative 3: Yamaha YZF 1000 R Alternative 4: Kawasaki Ninja ZX-10R
In the first stage of the application part, the criteria to be considered in the selection of motorcycles were determined as a result of interviews with motorcycle enthusiasts. Criteria explanations can be seen in Table 10. Motorcycle Selection The meanings of the evaluation criteria given in Table 9 are explained as follows: The criterion coded K1 is called "Engine performance". The measurement of engine performance, which directly affects adrenaline-generating events such as 0-100 and 0-200 acceleration, top speed, desired by sports motorcycle enthusiasts, was evaluated in kilowatt in this study. The numerical magnitude of the engine power value will have a positive effect on take-off acceleration, intermediate accelerations and top speed achievable.
The "Weight of the motorcycle" criterion is the K2 coded criterion. The weight criterion has a huge impact on acceleration, such as engine performance. Motorcycles with the same power will always accelerate in less time with less weight. In addition, the increase in weight will negatively affect fuel consumption. The low weight also has a positive effect on handling. Finally, outside of driving; we can say that light motorcycles create a better user experience when parking and in subtle manoeuvres.
The K3 coded criterion is the "Fuel consumption" criterion. In the study, the average fuel value consumed in 100 kilometres given by the manufacturers was used. Users expect low fuel consumption; It can be said that among the engines that produce the same power, the one with less fuel consumption is more efficient, which is a situation that motorcycle enthusiasts can enjoy.
The K4 coded criterion is the "Price" criterion. We can think that using a sports motorcycle, which is an expensive hobby, is considered by many motorcycle enthusiasts as price / performance. The same is true for automobiles. Considering that many evaluation magazines set an F/P criterion while scoring new models, we can say that the same is true for motorcycle enthusiasts. Therefore, potential customers are more interested in low-priced products.
The K5 coded criterion is "Design". This criterion is a more subjective criterion. In this study, the design criteria were found with their average values as a result of the interviews with the enthusiasts. The design of sports motorcycles, which can be considered a luxury hobby, is very important for sports enthusiasts. Harder and fiercer designs, bright colours and rim design affect the emotional reactions of the users and change the perception of quality.
The criterion coded K6 is "Fuel depot size". Considering that sports motorcycles burn a little more, they want the tank size to be sufficient for users to avoid the trouble of intercity trips and the trouble of constantly visiting the gas station.
The PIPRECIA method was used to investigate the effect of the criteria on the problem. After applying the PIPRECIA method steps, the results are given in Table 11. We can guess that the first thing that sports bikers consider is pleasure. The most important thing that affects the perception of pleasure in enthusiasts is acceleration and the feeling of acceleration. The two criteria that most affect acceleration, the K1 (motor performance) criterion with a score of 0,2076 and the K2 (weight) criterion with a score of 0,1977, had the highest weight in the overall evaluation result.
The weight values of fuel consumption, price and design are relatively close to each other. It can be said that each of these is included as a separate evaluation criterion under the expectation of acceleration and pleasure.
The fuel depot size criterion took in the last place with a score of 0,1093. The fact that this criterion, which has a more practical meaning when considering the desire of motorcycle enthusiasts, is in the last place shows that sports motorcycle enthusiasts put such criteria in the last place.
After the criterion weights were found, the initial decision matrix of the motorcycle alternatives was created. The decision matrix forming the initial step for all MCDM methods is given in Table 12. The values and rankings according to the SAW, WPM and ROV methods are shown in Table  14. After the calculations were made according to six different methods, Copeland processes were started in order to combine the obtained rankings. Comparative overall superiorities between alternatives are shown in Table 15.    Suzuki  3  0  3  1  Honda  1  -2  -1  3  Yamaha  2  -1  1  2  Kawasaki  0  -3  -3  4 Looking at Table 12, Suzuki's model came to the fore as the best alternative in the ranking seen in Table 17, although it was ranked 2nd according to the "engine performance" criterion and 3rd according to the "weight" criterion. Suzuki, which is close to the models in the first row according to the counterweight criterion, has a much cheaper price tag than the Honda with the best engine performance. Suzuki also came to the fore in the "design" criterion and became the most admired model with 8,5 points. The differences in these areas overshadowed the low scores in fuel consumption and tank size, and the overall assessment was placed in the first place. Kawasaki's model took the last place in this evaluation. Although it offers the best data on fuel consumption, the worst engine power and weight values influencing acceleration among these four and the second highest price tag could not save Kawasaki from being the last alternative.

Conclusion
Motorcycle selection can lead to indecision for sport motorcycle fans. Very similar offerings by competing firms and analysis of all technical details can be considered as causes for this hard decision problem. However, being a niche market and users being mostly hobbyist or adventurous people will push some criteria forward as being more important for these users. These are motor performance and weights of the motorcycles, which affects the most pleasure and feel of driving and acceleration. These two leave other criteria behind, such as fuel tank size which is more of a practical issue. Also, all alternatives in this study have very similar sizes of fuel tank and weights. It would be critical for firms trying to have an edge, to improve themselves in the areas of motor power while keeping a close eye on fuel consumption and to create a stylish aesthetic design. The prices will always be important in customer decisions as well.
In this study, results from different MCDM methods applied to a motorcycle selection problem dataset are combined with the Copeland method. In future studies, MCDM methods which are used in this study or other methods can be used on another problem and dataset, which then can be combined with the Copeland method for further analysis. This study will be helpful to sport motorcycle fans and producers in the market and may offer new insights.